DOCUMENT RESUME

ED 262 540 EC 180 948

TITLE Invitational Research Symposium on Special EducationTechnology, Symposium Proceedings (Reston, Virginia,June 2-4, 1985).

INSTITUTION Council for Exceptional Children, Reston, VA. Centerfor Special Education Technology.

SPONS AGENCY Special Education Programs (ED/OSERS), Washington,Dr,

PUB DATE 85NOTE 167p.PUB TYPE Collected Works - Conference Proceedings (021)

EDRS PRICE MF01/PC07 Plus Postage.DESCRIPTORS Computer Assisted Instruction; Computer Oriented

Programs; *Disabilities; *Educational Technology;Federal Aid; *Microcomputers; *Research; *SpecialEducation

ABSTRACTProceedings of the 1985 Invitational Research

Symposium on Special Education Technology are summarized. Thedocument contains a summary of each of 28 individual presentations,plus 10 research summaries submitted by individual participants, and42 summaries on current technology-based research in specialeducation. The first section reviews presentations in seven sessions.(1) rationale for technology research, (2) research on theeffectiveness of microcomputers in special education, (3) researchand development of special education authoring systems, (4) majortechnology-based research initiatives, (5) the use of advancedtechnologies in special education, (6) reports of research inprogress, and (7) funding for technology-based research. Researchsummaries review procedures and findings and list the funding sourceand the researcher's name and address. (CL)

***********************************************************************Reproductions supplied by EDRS are the best that can be made

from the original document.***********************************************************************

U.S. DEPARTMENT OF EDUCATIONNATIONAL INSTITUTE OF EDUCATION

EDUCATIONAL RESOURCES INFORMATIONCENTER IERICI

This document has been reproduced asreceived horn the person or organizationoriginating itMinor changes have been made to improve

reproduction quality

Points of now or opinions stated in this decoment do not necessarily represent office! NIE

position or policy

.INVITATIONAL

ResearchON

SPECIALSPECIAL EDUCATIONTECHNOIDGY

SYMPOSIUM PROCEEDINGS

"N) Conducted byL\ The Center for Special Education Technology

C.) Funded by

1-1-1Office of Special Education ProgramsU.S Department of Education

June 2-4, 1985

2

TABLE OF CONTENTS

I. Preface

Page

1

II. Summary of Presentations 2

A. Session 1: Technology Research: WhyBother? 3

Martin J. Kaufman 4

Ted Hasselbring 6

John Alden 8

James Kulik 10Kathleen M. Hurley 12Robert Yin 14

B. Session 2: Research on the Effectivenessof Microcomputers in Special Education

Herbert RiethMelvyn Semmel

161719

David Malouf 22

C. Session 3: Research and Development ofSpecial Education Authoring Systems 25

Joseph P. Lamos 26Marion Panyan and Paul Hazan 28Robert A. Zuckerman 30

D. Session 4: Major Technology-Based ResearchInitiatives 33

Gregory Jackson 34

E. Session 5: The Use of Advanced Technologiesin Special Education 37

Gwendolyn B. Moore and Elizabeth A. Lahm 38

F. Session 6: Reports of Research in Progress 41Susan Jo Russell 42Raymond G. Romanczyk 44Richard P. Brinker 46

Richard b. Howell 48Ron Thorkildsen 50

Philip M. Prinz 52Roger Awad-Edwards 54Al Cavalier 56

G. Session 7: Funding for Technology-BasedResearch 58

Mary K. Leonard 59J. Richard Taft 60Edward Esty 61

David Gray 61Richard Johnson 62Martin J. Kaufman 62

III. Appendix: Research Summaries 64

PREFACE

The 1985 Invitational Research Symposium on SpecialEducation Technology was conducted by the Center for SpecialEducation Technology. The Center, funded by the Office ofSpecial Education Programs of the U.S. Department of Education,is a national resource for information concerning the use oftechnology in the education of handicapped children and youth.

The purpose of the symposium was to provide a forum forresearchers actively involved in the systematic investigation oftechnology in special education to:

examine the status of technology-based research and theissues encountered by investigators;

exchange information about current special educationtechnology research projects; and

develop formal and informal mechanisms for networkingamong active researchers.

The symposium program included seven major sessions andtwenty-eight individual presentations. In addition to thepresentations, participants attending the meeting provided abrief summary of current research investigations. Thisproceedings document contains a summary of each presentation andthe research summaries submitted by individual participants.There are three sections in the document.

Section 1: Presentation SummariesSection 2: Research Summaries: PresentersSection 3: Research Summaries: Participants

The twenty-eight presentation summaries in Section 1 were pre-pared by Center staff based on taped recordings of each session.The ten research summaries in Section 2 were written by theindividual investigators and complement the material in Section1. The forty-two summaries in the last section were submitted byparticipants and represent an information base on current tech-nology-based research in special education.

The Center for Special Education TechnologyThe Council for Exceptional Children

1920 Association DriveReston, Virginia 22091

703-620-3660

1 4

SUMMARY OF PRESENTATIONS

0

SUMMARIES ARE ORDERED AS PRESENTED

Martin KaufmanTed Hasselbring

John AldenJames A. Kulik

Kathleen M. HurleyRobert Yin

Herbert RiethMelvyn I. Semmel

David MaloufJoseph Lamos

Marion Panyan/Paul HazanRobert ZuckermanGregory Jackson

Gwendolyn Moore/Liz LahmSusan Jo RussellRaymond Romanczyk

Richard P. BrinkerRichard D. HowellRon ThorkildsenPhilip M. Prinz

Roger Awad-EdwardsAl Cavalier

Mary LeonardRichard TaftEdward EstyDavid Gray

Richard JohnsonMartin Kaufman

6

2

SESSION 1: TECHNOLOGY RESEARCH: WHY BOTHER?

Martin J. Kaufman, Director of the Division ofEducational Services, SEP/OSERS, welcomed the symposium

participants and discussed SEP's commitment to educationalresearch.

The purpose of the opening panel presentation was to offer arange of perspectives on the status of technology-based

research. Speakers, representing commercial and academicperspectives, identified issues faced by researchers,research questions that need to be investigated, and

alternate research designs for the systematic investigationof educational technology.

Ted Hasselbring, Peabody College, Vanderbilt UniversityJohn Alden, Texas Instruments

James Kulik, The University of MichiganKathleen Hurley, Grolier Electronic Publishing

Robert Yin, The COSMOS Corporation

KAUFMAN

Dr. Martin Kaufman, Director, Division of Educational Ser-vices, OSEP, welcomed the participants to the symposium.

Dr. Kaufman summarized the mission of the Office of SpecialEducation Programs to provide leadership and fiscal resources forthe initiation, expansion and improvement of education servicesto handicapped children. The programs administered by OSEP areboth entitlement programs and discretionary programs. Withinthose two broad categories there are opportunities to support-research, training, demonstration, and technical assistance pro-jects. OSEP also provides policy clarification and compliancemonitoring. It is in the synergism of programs that a fasterrate of adoption for research programs is realized. The inte-grated set of comprehensive strategies provides a means forleadership and to flow fiscal resources to the special educationcommunity.

The systematic effect of those research programs is thatinnovative change occurs faster and more evenly on a nationalbasis than it would otherwise. The contributions of the researchprogram in general and, the technology research in particular,are in a number of areas. One contribution is in the area ofproviding new knowledge and understanding. Foi example, thetechnology effectiveness grants provide new knowledge, confirm

SOold knowledge and apply findings to the instructional design ofsoftware.

a)

A second contribution of the research program is to developand verify effective practice. The research program also contri-butes to innovative change by reviewing and analyzing currentresearch and practices. An example of fostering innovativechange is the research of COSMOS Corporation. That project isreviewing and analyzing current applications of and future pro-jections for artificial intelligence, robotics and simulation forhandicapped children.

A fourth area where research contributes is in the develop-ment of new or improved approaches to product development basedon research findings. Two examples in the technology program arethe efforts to develop authoring systems and the compensatoryapplications of technology. In these projects the concern is notonly to develop a product but to demonstrate the effective newdesigns and features of the product.

21 fifth area where the research program contributes is ininformation exchange. This happens in a number of ways. TheCenter for Special Education Technology makes information avail-able not only to the research community but to direct serviceproviders as well as institutions of higher education. Throughaccess to synthesized and current listings of information, hope-

fully the change process has more continuity as a result of lowercost access to information that is difficult to obtain andrapidly changing.

Finally, the research program provides the nation a problemsolving capacity. Active researchers represent a capacity totake experience, training, and knowledge, and apply it to emer-ging situations.

Research contributes to the Office's mission of initiating,expanding and improving services to handicapped children in seve-ral ways. The impact does not occur from a single study, becausea single study rarely is compelling, comprehensive and convincingenough in its findings to bring about change by itself. But itis through the cumulative effect of studies that a concept islegitimized and that innovative change begins. Through research,utility and effectiveness of concepts are demonstrated. Know-ledge and practice are incorporated into personnel preparationprograms. Findings are adopted as instructional design standardsby commercial nublishers. In the end, the contributions ofresearch do make a difference and they do bring about change.

The expectations for this symposium are concerned withinformation exchange and with the problem solving capacity of theparticipants. This meeting will enhance the continuity ofresearch and development. The program is designed to strenghtenthe performing community.

The research community, if it is to effect change, must bedisciplined enough to think in terms of how their researchaffects the school's ethos, the school organization, teachervalues and expectancies, staffing of schools, and curricula inte-gration. Only when we can state our visions and have our workfit those visions, can we integrate technology into the schoolsand classrooms for educating handicapped children.

HASSELBRING

Ted Hasselbring, Associate Professor at Peabody College,Vanderbilt University focused on the use of single subjectdesigns in technology-based research. He suggested a need tomove from attempts to validate some generalized effectivenessmodel of technology in large group studies to investigations thatfocus on the individual learner. He notes that others haveidentified other directions for technology-based research andsuggested investigating technology as it relates to individual-ized learning.

Hasselbring used data from a set of studies to illustratethe importance of analyzing individual data to assess the effec-tiveness of computer-based learning. The set of studies wasdesigned to examine the effects of arcade-style drill and prac-tice games on learning of mathematics disabled children.

One reason for choosing the arcade games format was the highreported use of that software by teachers. Surveys of softwareuse in the public schools suggested that in 80 percent or more ofcases, teachers were advocating the use of this type of drill andpractice software for mathematics instruction in special educa-tion. Another reason was that there was very little evidence ofthe effectiveness of that type of software even though it waswidely advocated.

The studies did not compare computerized drill and practiceagainst more traditional methods but looked within the drill andpractice format. Independent variables included: the speed ofthe game, the level of user control, the amount of play time, andthe length of time the game was used. The dependent variable inall cases was the subject's rate of correct responding.

Rate rather than accuracy was used because all subjectsscored very high on the pretest. Most were close to 100% accur-acy level. However, in terms of rates of addition and subtrac-tion, scores dropped drastically. The three studies had veryconsistent results. In virtually all cases the drill and prac-tice had some moderate positive effect when group data was ana-lyzed. Regardless of the variables manipulated, there were posi-tive gains in almost all cases. A consistent finding withingroups was that some subjects had an increase of over 100% in therate of correct responding and others made virtually no gains atall. Within groups gains ranged from low to high. Regardless ofthe treatment, group gains remained positive.

There is significant literature in cognitive psychology onmathematics, primarily with non-handicapped. One frequentlycited study by Groen and Parkman, looked at how students process2 digit addition problems. They proposed three cognitive modelsthat students use when processing addition problems: a counting-all model, a counting-on model and a min-model. The number ofincrements used to process addition problems varies with each

model. For example, if the student is using the counting-allmodel with the problem 3+5, the counter is set to 0 and theybegin to count. There are 8 increments to be counted using thismodel. The counting-on model differs in that the student setsthe counter to the first number encountered. In the case of 3+5,the counter is set to 3 and incremented 5 times. In the min-model the student identifies the larger number, in this case 5,and increments three times.

We applied these findings to our sample to see if studentswere using the min-model and also whether computerized drill andpractice would move them from a counting model, the min-model, toa reproductive process where they reproduced the answer frommemory. We took pretest data, latency of responding data on allsubjects and then provided drill and practice using arcade-stylesoftware. We found that after 20 to 30 days of computerizeddrill and practice students were still using a min-model. Atbest we were reducing the amount of time or making them moreefficient at using the min-model. That's why when looking atrates we got moderately positive gains. We also found that thereis a ceiling. There is a peak rate of responding when using themin-model and over time subjects who reached the ceiling beganmaking errors.

We concluded that drill and practice itself had very littleeffect on changing cognitive strategies the children used. So ifwe are trying to get students to memorize facts, the data sug-gests that computerized drill and practice is not very effectivein changing cognitive strategies. We also found that with arelatively short pretest we can determine what model an indivi-dual is using. The next step in this research is to intervene insome meaningful way that moves students from primative counting-on strategies to reproductive strategies. Our findings may beincorporated in software that identifies cognitive strategies andprovides effective interventions.

ALDEN

John Alden, Manager of Educational Marketing at TexasInstruments reviewed current trends in computer technology, iden-tified problems with current technology-based research in educa-tion and suggested some solutions to these problems.

Alden indicated that in fiscal year 1984, Texas Instrumentsspent 691 million dollars on research or about 10 percent of thecompany's gross sales of 5.7 billion. The forecast for 1985 is830 million dollars. The major gains of this research and deve-lopment are focused on productivity as well as on finding newproducts and services. The gains in productivity result frominnovation.

The changes in computer technology will impact educationalresearch investigations. Texas Instruments uses a term - AEGs -to measure the number of intelligent ac_ivities on a given chip.In a few years, there will be chips available that have a millionAEGs. As chips become more dense, the probability of error growsexponentially. The complexity of chips is at the point where itis impossible for a human being to control the design process.The point is that our ability to put information in a computerwill grow exponentially in the next few years.

Alden cited a recent National Academy of Sciences report onresearch in precollegiate education that suggests that most edu-cational research is behind the cutting edge of technology. Thereport noted three levels of computer use in education. Level 1is characterized as common uses of technology such as drill andpractice or programming. A level 2 application is defined asrepresenting a diversity of efforts that are important for dis-covering the ways to use technology effectively in education.Level 3 is characterized as advanced theory and systems develop-ment. An example is artificial intelligence (AI) research.Pragmatically the research is oriented toward constructing expertsystems that represent various knowledges both of the domain andof the learner.

Alden drew several conclusions about the status of educa-tional research. First, the current investment in micros haslimited effectiveness because the technology does not deal withhigher order thinking skills that are the basis for self directedlearning. Second, current practice in education has a disasso-ciation between formal schooling and experiential use learning.Third, the presence of micros has set the stage for a persuasivekind of computer literacy. The ubiquity of computers in theschools is raising the cumulative learning capacity of students.This increase in the overall learning capacity of students issomething very profound.

Fourth, symbolic computing or AI computing is the key foreducation research. It has a close tie to cognitive psychology.For example, LISP as a language is powerful and allows for the

8

12

quick prototyping of problems and programs. Research can alreadydemonstrate intelligent diagnostic coaching and discoverylearning systems. Increased scientific research will allow us totailor instruction for individuals to a degree not generally doneoutside of the apprenticeship that is prevalent in graduateschools and other research environments. New forms of AIresearch, especially from cognitive psychology, set the stage forhigher potential research benefits.

There are several things that can be done. First, all of usinvolved in research need to change our attitudes about whatresearch is. Alden suggests a need to change the metaphor onwhich research is based. An example is the Idea Program at TI.This program provides small grants to spur innovation well aheadof major funding in the laboratory. A good number of TI's bestinnovations, include Speak and Spell, came out of the Idea Pro-gram. The Program grants must lead to revolutionary improvementsin products or methods. The innovation should have step-functionimpact on some portion of TIs business. Alden suggested thenotion as a yardstick for R&D in education.

A second thing is to think about productivity. If you thinkabout productivity, you change the metaphor of research to con-centrate on pedagogical and curricular changes that result fromthe widespread use of computational power. Alden indicates thatfew examples of advanced research can be found in the educationalliterature -- there are no step-function studies. Alden indi-cated a need to concentrate on the level 3 kind of research. Hesaid that to do otherwise puts researchers far behind the leadingedge of technology.

9 1:3

KULIK

James A. Kulik of The University of Michigan reviewed thefindings of meta-analyses on computer-based education (CBE) con-ducted by investigators at the Center for Research on Learningand Teaching. Separate meta-analyses have been completed on theeffectiveness of computer-based education (CBE) in elementaryschools, high schools, colleges and nontraditional post secondaryinstitutions. Kulik defined meta-analysis as the statisticalanalysis of the results of a large collection of independentlyconducted studies. The meta-analyst uses subjective techniquesto search literature and library databases to find all thestudies done on a certain topic. All the features of the studiesare coded and expressed in quantifiable or comparable terms. Thefeatures coded are then statistically integrated to determinewhat study features are affecting the results, what the averageresults are, etc.

The purpose of the meta-analyses was to get an overview ofwhat the effects of CBE have been in the typical study. Onehundred ninety-nine formal controlled evaluations of the effectsof the computer in education were identified. Features of thestudies such as methodology used, the setting of study, thepublication source and history, as well as outcomes such aseffects on learning and attitudes were coded. The two sources ofdata were then analyzed to determine findings of the typicalstudy and whether certain types of studies were producing uniqueresults.

Kulik reported that the criterion in virtually all studieswas student learning. Overall results indicated that most CBEprograms have had positive effects on student learning. Specificfindings include: students have generally learned more inclasses when they have recieved help from computers; studentsremembered what they learned longer; students also learned theirlessons with less instructional time; students liked theirclasses more when they received computer help; and studentsdeveloped more positive attitudes toward computers when they asedthem in school. However, computers did not have positive effectsin every area studied.

CBE was not uniformly successful in all its guises and atall instructional levels. Effects of computer assisted instruc-tion (CAI) , computer-managed instruction (CMI), and computer-enriched instruction (CEI) were different at different instruc-tional levels. CAI was defined as drill and practice andtutorial instruction where students received instruction directlyfrom the computer. CMI included recordkeeping, online and off-line testing, and scheduling where students weren't directlyinstructed by the computer. CEI included uses of the computeras a tool for calculation, simulation, and programming.

Elementary schools have had a good deal of success in

increasing student achievement through CAI programs; they have

lo 14

had less demonstrable success with CMI and CEI. Secondaryschools have had success with CAI and CMI but less success withCEI. College and adult courses have benefited moderately fromCAI, CMI, and CEI. Future programs of implementation anddevelopment should also take these findings into account.

Both journal articles and dissertations present a basicallypositive picture of CBE effectiveness, but the findings reportedin journal articles are clearly more favorable. Research is

badly needed to determine the factors that produce differences injournal and dissertation results because such differences havebeen found in a number of different areas in social scienceresearch. Does editorial gatekeeping lead professional journalsto present a distorted picture of social science findings? Or dodissertation authors simply measure experimental effects lesswell than do more seasoned researchers?

Although a variety of different research designs can be usedto show the effectiveness of CBE, certain research designs seemto produce more positive results. Studies where the sameinstructor teaches both experimental and control classes, forexample, report somewhat weaker effects than do studies withdifferent experimental and control teachers. Studies of longduration often report weaker effects than do short studies.Reasons for the difference in results from studies using dif-ferent experimental designs are imperfectly understood, however.Research on such factors should be encouraged.

HURLEY

Kathleen Hurley, Vice-President of Microcomputer Softwareat Grolier Electronic Publishing, focused on how publishers canwork with researchers. Hurley indicated that for publishers it'sa time when production of educational materials is expensive.Publishers find themselves spending $50,000 to $100,000 for a

product that is dated as soon as it is published. From the

viewpoint of a commercial publisher, the first consideration is

one of dollars and cents. Hurley indicated that many think theeducational software market is very good. In 1983, 93 milliondollars was spent on educational software for use in the home andin schools. In 1984, 186 million dollars was spent. It is

projected that 318 million dollars will be spent in 1985.

In 1983, the schools spent 38 million dollars on software.However software sales represented only 3.6% of the total elemen-tary and high school instructional materials sales. In dollarsthis amounts to 38 million for software versus a total of almost2 billion for other kinds of instructional material. The specialeducation market represents only about 10% of the overall popula-tion in schools. Assuming that sales of 5000 copies of a productis considered satisfactory by an educational publisher, in thespecial education market the projected sales figure would bereduced to 500 copies. The average product cost $30,000 todevelop. The retail cost of the product is about $60.00. Add tothat the manufacturing, advertising and marketing costs and thepublisher's cost per unit on sales of 500 units becomes prohibi-tive.

Hurley suggests that there are several solutions to theproblem of developing materials for special education if publi-shers and researchers work together. One suggestion is thatthere be subsidies or shared costs from the Department of Educa-tion or educational associations that would provide incentivesfor publishers to develop and market special education software.Breaks for publishers on advertising in special education jour-nals would make their products known in the special educationmarketplace. Hurley also suggested collaboration between publi-shers and educators to adapt or modify existing software for usewith both handicapped and non-handicapped populations. Thiswould ensure lower cost per unit through higher production.

Another suggestion is that publishers make more use of

research findings. For example, publishers could call on resear-chers to identify and interpret data that relates to productdevelopment. Research findings could be incorporated into pro-duct planning and specifications. Publishers could field testproducts using accepted techniques derived from the work of

researchers. They could also revise existing software programsusing data from current research efforts.

Publishers want to know specific considerations to be awareof, such as the presence or absence of graphics, when developing

-1216

special education software. They want to know what adaptionshould be incorporated in a product for different handicappingconditions or what field testing techniques can be used and howmuch field testing is necessary.

Hurley indicated that software publishers have becomeincreasingly more sophisticated regarding the development ofspecial education software. However, there are several issuesthat publishers continue to face in order to keep pace withsoftware developments. The first issue is a dollars and centsone. Questions of a viable market, recovery of developmentcosts, effective entry into the special education market andcollaboration with researchers and other education professionalspersist. Second, is the inability of publishers to keep abreastof the current research and to find ways to collaborate withresearchers in the design of special education software. Guidelines need to be developed that will result in simple, consistentprinted commands, minimal lines of print, and other relevantdesign features. A related issue is how to go from theory topractice. Publishers must be able to interpret research oncomputer technology and use it to develop practical software.There is a lot of descriptive research available, but more basicand applied research that will contribute to product developmentis needed. Finally, collaboration between publishers and researchers is needed. Ideally, there would be shared costs forinitial conceptualization and field testing. Closer collaboration between publishers, the Department of Education, and CEC andother professional organizations is also needed. The last areaof collaboration is that of researchers and hardware manufacturers to plan low cost adaptions and modifications to computersfor use with handicapped students.

YIN

Dr. Robert Yin, President of COSMOS Corporation, indicatedthat researchers dealing with rapidly advancing technology haveto identify the limitations and possible adaptations of theresearch methods used. The case study is a powerful tool fortechnolow:based research. Yin listed five topics to considerwhen using this methodology: defining the problem, designing thestudy, collecting the data, analyzing the data, and reporting thefindings. The literature on research methodology addresses datacollection problems but provides little information on thedefinitional or design problems.

The consummate characteristic of a case study is the abilityto study the context and not just the phenomena. Sometimes thephenomenon studied does not have sharp boundaries from the con-text. Inevitably there are more variables than data points. Itis a different technical situation. You can't talk about var-iance and means, you must search for other techniques. Inorganizational studies, implementation studies, or studies of thedevelopment of technology, it is difficult to limit the study toa particular phenomenon resulting in analytic problems.

Yin discussed COSMOS' study of advanced technology to illus-trate the considerations in defining a case. The study was toidentify existing uses of robotics, artificial intelligence (AI)and simulation outside of education and to determine possiblefuture applications for special education. The problem from theresearch point of view was to define what was to be studied. Inrobotics for example, the first inclination was to study a robotin the lab or an industrial setting. What is generally found inthese settings is an arm, a gripper, a computer and an environ-ment that provide things to be manipulated. That is the mostcommon application of robots today. If the researcher haddefined the case as this apparatus, subsequent studies wouldprobably focus on the technical development of the apparatus. Toget a potential use, a more abstract unit of analysis called anapplication was developed. The apparatus under study had to bein use on a routine basis in a real environment. That wholesituation would be the subject of study. If the original casehad been defined as the technology, it would have led to one setof inquiries. If the case was defined as an application, therewould be an entirely different set of inquiries. It is not aquestion of one definition being right or wrong, but the impor-tance of the choice a researcher makes at the beginning of thecase study.

Yin sited another example of a study of microcomputer imple-mentation in special education programs. The focus was on imple-mentation so effects were not studied. The question asked was --why in some schools, in some programs,is microcomputer usegrowing so rapidly? We wanted to examine different factors suchas wealth of the district, teacher training, etc. however, onlytwelve districts could be included in the study. This is a

.,14

18

design problem -- how do you choose 12 districts when you want torepresent district size, region of the country, racial, ethic,income and curriculum variables? Twelve cases is not enough.

To choose 12 sites as a good paradigm is the same thing aschoosing 12 experiments in the lab. What you have to do isdevelop some theories. You look for situations in which you canreplicate, even if you can't manipulate. If you can predict thatthe same things will happen in two, three or four cases and theydo replicate each other, that's a very potent outcome. Twelvebecomes a large number. We can choose four cases because theyconcentrated on administrative use only. Our prediction is thatif you concentrate only on adminstrative uses the system will notgrow. It will rapidly become a mainframe system. You can chooseanother four cases because they only use the microcomputers forinstructional applications. The prediction is that use will growbut not as rapidly as the third situation, which is when thereare computers for both administration and instructional applica-tions. The prediction is that in those districts there is acomplementarity of applications that draw the best organizationalsupport. We predicted better, smoother implementation in thosedistricts.

If you can think about 12 being a large number in thatsituation, a lot of things can be accomplished with that kind oflogic and that kind of design. Yin's examples illustrated two ofthe more difficult steps in conducting case study research. Casestudy research is an appropriate technique for studying tech-nology. With careful attention to definitional and design prob-lems, case study methodology can be rigorously applied.

1915

SESSION 2: RESEARCH ON THE EFFECTIVENESS OF MICROCOMPUTERSIN SPECIAL EDUCATION

Reports of three major SEP-funded studies on the use andeffectiveness of microcomputers with mildly

handicapped students. The four year studies are intended tobuild an initial knowledge base on effective computer use

in special education. The principal investigators discussedresearch design and methodology, issues encountered in conducting

the research investigation and the preliminary findingsof their investigations.

Herbert Rieth, Indiana UniversityMelvyn Semmel, The University of California at Santa Barbara

David Malouf, The University of Maryland

16 20

RIETH

Dr. Rieth, of Indiana University, reviewed findings fromearlier research at The Center for Innovation in Teaching theHandicapped (CITH), that had bearing on the current investigationof microcomptuer effectiveness. In an earlier study of secondaryenvironments, teachers designed classroom environments usingprinciples of effective teaching. The secondary environmentswere effective in producing academic gains. Mildly handicapped,inner city students gained two and three years. There was alsosome data to suggest that microcomputers could have a deliriouseffect on student performance. That is, the effect of goodinstruction could be neutralized by ineffective use of micros.The data suggested that students were not necessarily learningeven when micros were used extensively and were viewed byteachers as an integral part of the classroom instructionalenvironments. Findings suggested that teachers weren't lookingat the assessment information provided by the computer. As aresult students using drill and practice software were oftenworking on skills learned two or three years earlier. Anotherfinding was that even though students were not gaining academi-cally, they were still attending to the task at a very high rate.The rates of engagement on the current study are about 90%. Thisis active engagement where students are emitting responses.

In most of the general teacher effectiveness literature,active engagement is usually correlated with achievement growth.The current research suggests that to some extent the microcompu-ter magnified many of the instructional problems in classrooms.Teachers are ineffective in monitoring students. There are casesof students making the same errors over a two week period andteachers were not using the data available to detect theseerrors.

The present tesearch is being conducted in three schoolsystems -- a large urban district, a suburban and a semi-ruralschool district. In the first year of the study quantitative,and qualitative methodologies were used to study the generalcontext of the secondary school community and the impact ofmicrocomputers in that environment. Impact on student behavior,teacher behavior and curricular content were studied, as were therange of materials used for instruction, the frequency of use ofdifferent types of materials, and grouping practices in secondaryclassrooms. Classroom observations and in :erviews with adminis-trators, teachers, parents and students were conducted to providedescriptive information on general ecology of the secondaryenvironment and the use of microcomputers.

First year data were collected using an observation systemdeveloped in an earlier study of academic learning time. Tech-nology was an integral part of the data collection process. Thewhole process was paperless using Epson HX20's and a communica-tion package for data transfer. During the first year over18,000 observations were conducted and no data was lost. A point

17 21

time sampling system was used. Students were randomly selectedfor observation. Behavior of target students was coded asactively engaged, passively engaged, engaged with directions, andoff task. Teacher behavior was also observed.

Findings indicated that the highest percentage of computeruse was in the content area of mathematics. Computers in class-rooms were in use about twenty-five percent of the time. Theaverage time for individual students was 12 to 13 minutes. Much ofthe computer-based instruction occured in conjuction with paper-and-pencil activities. Reith reported that student computer usealso freed up teacher time. However, that time was not reinves-ted in instruction, but rather in supervision-type activities.

Student engagement rates across the three types of engage-ment was close to 80%. Students were operating in isolation.The amount of academic feedback was only 1 to 2%. This findingled to an intervention study to determine how feedback can bedelivered to students engaged in computer learning.

Researchers found that feedback was effective. The longerstudents were on the computer, and the more feedback they got,the better their academic performance. However, teachers couldnot provide feedback to students using the computer and maintainfeedback to other students as well. Profiles of teacher behaviorwere roughly comparable when the computers in use and when it'snot in use. This suggests that teachers are not doing muchdifferent when the computer is in use. The presence of thedevice does not alter teacher behavior, even though the districtsin the study provided training on the effective use of computerfor teachers. A training intervention was designed. Researchersfound that providing teachers with general strategies for inte-grating microcomputers into the classrooms was not effective as atraining strategy. Training teachers to use a specific piece ofsoftware was more effective.

The third and fourth years of the study will extend thepreliminary efforts at intervention using applied behavioranalysis methodologies. Environments for teacher training andeffective use of software will be designed and the impact onlearning outcomes measured.

SEMMEL

Dr. Semmel, of the University of California at SantaBarbara, discussed the approach used in the four-year microcom-puter effectiveness study and th? theoretical notions emergingfrom that approach. One of the problems at the outset of thestudy was to conceptualize the aspects of microcomputer use to bestudied. A theoretical framework to guide the research investi-gation was developed. The problems in assessing the effective-ness of microcomputers with mildly handicapped elementary schoolyoungsters were viewed in terms of three facets in a quasi-pathmodel. The three facets were defined as allocation access vari-abl'7, and use and outcome variables. A comprehensive system wasnecessary because of the likelihood that these are exogenousforces or causal constraints on each subsequent level of theproblem. For example, the process by which technology is allo-cated has an impact on the school site distribution of tech-nology, which in turn has an impact on administrative arrangementand delivery system variables.

Researchers developed the concept of the Micro-EducationalEnvironments (MEEs) and looked for variation in learner charac-teristics, technical configurations, instructional content andcharacteristics of personnel involved in each environment.Combinations and variations of those variables resulted in varia-tions between MEEs. It is the MEEs that have a direct effect oncomputer access leading to pupil use and eventual effects.

The overall strategy for the four year study was to startwith outcome prediction studies with fidelity to standardsampling techniques allowing generalization to the populations ofinterest. Survey, into :view and direct observation techniqueswere used. The idea in the initial data collection phase was todevelop a number of sources of data on similar questions. In thesecond year, that data would be used to develop some hypothesesrelative to MEEs that might be effective, identifying the vari-ables and constructing different models in the laboratory. In

subsequent years, controlled applied studies and validationstudies will be conducted in the field with the eventual goal ofdeveloping a delivery system model.

Semmel reported on the first year of data collection fromthe administrative and teacher surveys and the student/classroomobservations. Also reported were the results of an independentethnographic study in two school districts using case studymethodology, in relation to quantitative findings. In thisstudy, a second research team is working blind with no access tothe results of other studies. This strategy permits the crossvalidation of qualitative and quantitative findings.

In looking at allocation variables, one question was todetermine the status of hardware in schools. A survey of a

random sample of administrators provided information about thedistribution of microcomputers. Apple computers were the domi-

nant hardware at the elementary level. In the study at Indiana,Rieth found few Apples at the secondary level. Studies will beconducted to assess the problems students have in moving from onetype of computer to another.

Fifty one percent of microcomputers used in special educa-tion are acquired through general funds. In these Californiadistricts sampled, only 8% of hardware was acquired throughdonations including the Apple give-away program. In general theacquisition of equipment was very rapid. One concern with thecurrent acceptance of 32 bit machines, particularly theMacintosh, is that scnools will be unable to reinvest in newequipment and will be forced to continue using the Ile. A ten-dency toward centralization of hardware and software acquisitiondecisions was identified. Centralization at the district levelis rather distal from where the microcomputer instruction is tobe delivered.

Teachers in special day classes, resource rooms and regulareducation (mainstream) settings were surveyed. Eighty-two per-cent of special class, 60% of resource room teachers and 91% ofregular class teachers reported having access to microcomputers.Forty percent of resource room teachers selected randomly did nothave access to microcomputers compared to only 8% of regularclass teachers. A mobile cart was the predominent mode of loca-ting equipment for all classrooms except the resource room. Thisfinding shows the general demand and shortage of hardware thatnecessitates transporting equipment. It also demonstrates adifferentiation with respect to regular and special class set-tings.

A relevant finding in the case studies was that schools gothrough a natural evolution with respect to the allocation ofhardware. The process begins informally with someone beinginterested in the technology. It evolves as the technology getsinto the schools. Competition for the equipment starts veryrapidly and leads 10 a distribution by class. That distributionby class causes conflict and eventually tremendous pressuresbuild up to centralize microcomputer equipment into a lab.Schools with established microcomputer delivery systems aremoving to centralization very rapidly. They are also centrali-zing the curriculum around the equipment.

The teacher survey data showed that group instruction is animportant delivery device as compared to individualized instruc-tion. This finding has implications for the development of MEEsand educational software. Teachers in all classroom settingsindicated that the primary use of microcomputers was to supple-ment or reinforce existing programs. Use of computers to teachnew skills was rated lowest in special education settings. Thepredominant intention of the Leacher is not that the computerteach new skills.

A direct observation study of ninety classes validatedfindings about the predominancy of Apples and also showed more

"off- brand ", obsolete equipment being used in special dayclasses. This finding has implications with regard to the quan-tity and quality of software products available for this setting.In addition, it could be concluded that microcomputer equipmentis being segregated with students in segregated settings.Special education personnel are not taking advantage of theopportunity to integrate students with lab settings.

Levels of engagement were high in all settings. The atten-tional deficits ordinarily seen with mildly handicapped studentsare not observed when students are using the computer. Whenengaged in computer learning with peers, engagement in mainstreamsettings is relatively high. Group activity does not produce ahigh degree of off task behavior. Ethnographic interviews andobservations suggest that the high level of group use is a stricteconomic distribution of students and machines. The demand is sogreat that it is dictating grouping patterns.

The rate of teacher contact and monitoring of students whileon the computer is very low. Data from studies indicate veryclearly that error rates are high, above twenty percent. Thepossibility is that we are engineering students to attend to higherror rates. In the lab, students are responding impulsively.They are being reinforced by the continuous feedback. Instead ofprompting reflectivity, the software being used may be reinfor-cing impulsivity.

A research team is studying the implications of working ingroups. The kinds of relationships that develop from a communi-cations point of view have implications for the current conceptof engagement. It is likely that automaticity issues will bereduced if another student provides the correct resopnse in aproblem solving situation.

e

MALOUF

David Malouf, of the University of Maryland, described theresearch on the effectiveness of microcomputer technology beingconducted at the Institute for the Study of Exceptional Childrenand Youth. He gave an overview of the studies completed and thendiscussed research designs and issues encountered.

The studies of microcomputer effectiveness are being conduc-ted in a large suburban school district primarily at the elemen-tary level with some work at the middle and junior high level.Two descriptive studies have been completed. The first natural-istic study was to identify special educators in the district whowere using microcomputers and to conduct observations and inter-views. The second descriptive study was focused on the difficul-ties learning disabled students have in learning to use wordprocessing. Two additional descriptive studies are in progress.One is looking at successful strategies students use with compu-ters and the other is a study of students working at the computerin groups.

Several experimental studies were conducted to compare com-puter versus non-computer learning. A structured observationstudy compared computer drill and practice to seat work typedrill and practice. A study of perceived mastery compared thesame task presented on the computer and off the computer. Theimpact of computer assisted drill on achievement was a studycomparing drill and practice on and off the computer. The "com-posing stories" study looked at the processes and products ofwriting in one computer condition and two non-computer condi-tions. Computer versus computer studies compared computer gameswith drill and practice on subsequent student motivation. Alatency study compared retention on computer-based learningtasks.

Malouf reviewed some of the findings and issues identifiedby the researchers. He pointed out the difficulties in largescale naturalistic research that yield a large amount of data.He suggested as a research strategy, more focused descriptivestudies. For example, the study of students using word proces-sing resulted in rich information and a manageable amount ofdata. It provided information on those learning problems thatare amenable to training and those which are more persistent.The findings give direction for further descriptive and experi-mental studies.

A structured observation involved the same students workingon computerized drill and practice and paper-and-pencil drill andpractice. Three areas of student behavior were observed -- offtask, system and on task. Attention to system included activi-ties that were on task but had to do with the system rather thanthe content of the activity. Students were almost never off-taskwhen working on the computer. The findings on "system time" haveimplications for the time allocated for computer use, software

22 26

design and classroom procedures. A distinction should be madebetween time engaged in the mehanics of the system and the timeengaged in learning the content. While observing students,teacher interactions such as help with content, help with system,off task, behavior management, were monitored. Teacher inter-actions were higher with students using the computer than withstudents doing se-t work. There was also a greater amount of"system help" on the computer than on seat work.

One problem in conducting computer effectiveness research isthat of variable confounds. Findings reported may be confoundsrather than essential aspects of instructions. For example, inthe observations study, researchers did not intervene but therewere numerous variables operating in the classroom in addition tothe computer. The computer was often physically removed from thegroup of students reducing interaction with peers. The learningactivities presented on the computer tended to be of shorterduration than seatwork learning activities. The content oflessons also tended to be different. Not only was the computerphysically removed, it was also curricularly removed. Withvariable confounds operating, it is not possible to clearlyestablish the effects of computer-based learning. The informa-tion gained in the observational study was valuable descriptivelybut not for explanation.

The issue of variable confounds led to a study of computerassisted drill on achievement. Achievement, cognitive and affec-tive factors were studied with an attempt to control some of theconfounds. Control of experimental conditions is another issue.For example, in the computer condition students received imme-diate feedback. Feedback in the paper-and-pencil condition wasdelayed. One approach was to develop real counterparts in eachcondition such as providing immediate feedback in the paper-and-pencil condition. The study of computer games compared twoconditions. One group was engaged in computer games the othergroup in computer based drill and practice.

Researchers examined the relationship between responselatency and retention. There is ample justificaiton forrecording a time element when monitoring student learning on a

computer. In non-computer instruction the dimension mostcommonly used is rate. For computer-based learning researchersrecorded latency rather than rate. In a study where studentswere taught a functional reading task of matching words withtheir common abbreviations, accuracy and latency were recordedwith each trial. A long term retention test followed thetraining. Data was examined for systematic differences inlatency between words retained and those not retained. A con-struct, the trial of last error, from the literature on latency,is the trial upon which the subject makes the last error prior toreaching a criterion. The latency before and after the trial oflast error behaves very differently. Before the trial of lasterror the subject was still working on accuracy. On the trialafter reaching criterion, accuracy decreased systematically.This finding should improve the ability to predict whether a

,2327

student has learned the content.

An assumption in the survey research was that one of thecritical functions a teacher performs is in evaluating andselecting software. A survey of 400 teachers reports lessinterest in the evalution instruction than in information pro-vided by a colleague. In response to best methods for selectingsoftware, the preferred method was to try out the software.Reviewing software evaluations was rated low by teachers.

Malouf concluded that in research on technology in educationthere is a tendency to look for simple answers to complex ques-tions. He suggested that research was more likely to providecomplex answers to simple questions.

2824

SESSION 3: RESEARCH AND DEVELOPMENT OF SPECIAL EDUCATIONAUTHORING SYSTEMS

Reports of three SEP-funded contracts to developauthoring systems for use in special education. The eighteen

month projects are intended to provide tools for thedevelopment of microcomputer software for the individualized

instruction of handicapped children. The principal investigatorsdiscussed the design of authoring systems and the issues

encountered in development.

Joseph P. Lamos, Denver Research InstituteMarion Panyan and Paul Hazan, The Johns Hopkins University

Robert A. Zuckerman, Kent State University

LAMOS

Joseph Lamos of the University of Denver discussed thedevelopment of the H.E.L.P. Authoring System. Lamos indicatedthat his background as a systems analyst influenced the design ofthe educational authoring system. In 1970 Lamos was involved inthe development of a large computer-based instructional systemfor the military. The system was designed to provide indivi-dualized instruction to compensate for lack of basic skills.There were similarities between the problems faced in educatingthat adult population and the problems of the mildly handicappedin school settings. One similarity was the importance of lookingat the entire system including hardware and the cultural environ-ment in which the system is will be embedded. Adapting princi-ples of systems engineering, Lamos developed a notion of instruc-tional engineering that involved a synthesis of research andtheory in educational psychology and cognitive psychology. Partof that engineering perspective is that a functional analysis isnecessary to determine the features a system needs to operateproperly and to meet the requirements of the people using it.

The other aspect of development is operational analysis. How isthe system to be used and what are the constraints in theenvironment that may limit and inhibit the use of the product.

One goal was to design an authoring system that would movetoward the principle of integrating special education studentsinto the regular classroom. For that to happen, the product mustserve the needs of both special education and regular educationteachers working as team to meet the needs of children. Theother aspect was to focus on how the microcomputer was integratedinto the classroom situation. Data on computer use suggestedthat schools often begin with one machine in each classroom buteventually move to a centralized lab environment. That change inhardware location had to be considered in designing the system.Another aspect of the environment was that the classroom settingis activity oriented. That is, the teacher moves from activityto activity and the micro and the software had to fit into thatway of operating. Typically the student is engaged in somethingfor 15 minutes and then moves on to another task.

The authoring system was designed as a tool to allow peopleto adapt courseware to the learning needs of an individual stu-dent. Adaptation occurs at two levels, the micro-level in thesense of producing an instructional flow within the process ofteaching a concept. At the higher, curriculum ordering level, itis necessary to reorder larger chucks of instructional materialto adapt to individual learning requirements. An authoring sys-tem should have the flexibility to implement a variety ofinstructional strategies and also the capability to take a stra-tegy and apply it across a variety of content areas. The systemshould engage the teacher in student management without over-whelming detail. The developer also needs to focus on organiza-tions. Initially, schools should be able to use the system at avery basic level.

26 3 0

There are a number of ways to author CAI courseware. Indesigning an authoring system there is a tension between ease ofuse and flexibilty. The use of higher order language to designinstruction is time consuming and difficult. Authoring lan-guages offer another option. A modular authoring system has thecapability to create small units of instruction that can becombined and used in a number of ways. In the H.E.L.P. AuthoringSystem two levels are used. The first is micro-level authoring,where editors are used to structure the instructional strategyand design computer graphics.

At micro-level the idea is to create micro units, meant toprovide instruction on a single concept. It is at this levelthat adaptation to the needs of the special student comes intoplay. Parallel micro units can be developed in a content areafor different learning needs. The units are then collected inlibraries. A micro unit involves the student in 10-15 minutes ofinteractive learning. Within a micro-unit the major element is adisplay or single visual field. Within the display are somestandard presentation and and control items.

When designing an authoring system, there is a constanttension between the current and future needs of the organization.The H.E.L.P. system allows the schools to begin the authoringprocess at a basic level and to expand with their increasedunderstanding of the system.

27 31

PANYAN AND HAZAN

Marion Panyan presented an overview of the authoring systemdeveloped by the Division of Education and The Applied PhysicsLaboratory at The Johns Hopkins University. The purpose of theauthoring system is to provide teachers with a tool that could bewidely used to create lessons in discrimination, receptive lan-guage tasks, and basic reading and math skills. The systemconsists of 12 different lesson types. By varying items, num-bers, letters, and words, different lesson types can be created.Once the lesson type is chosen, the next decision for the teacheris the type of instruction. By using the same lesson type butvarying the text instruction, an entirely different objective isaddressed. Teachers are able to create lessons by selecting textand graphics stored in the system. There are currently 96graphics. An option for teachers to create additional graphicswill be incorporated in the software. In addition to selectingthe content of the lesson, there are important process and formatvariables to consider. Through a series of consecutive menusthese decisions can be made throughout the authoring process.Another feature of the system is the student recordkeepingfunction. Two types of data are collected for each student. Thefirst type provides information on accuracy of student perfor-mance. In pilot tests of the system, teachers have not foundthose data very useful because they are not graphically por-trayed. It does not give them enough direction to compose orconstruct the next lesson. The second type of data is a trial-by-trial analysis. These data help to determine a positionpreference or other error type. The program can then be adjustedaccordingly.

One of the principles guiding development of the system wasteacher acceptance. Representatives from eight counties in Mary-land and one in Virginia and Pennsylvania were invited to use thesystem and complete a field test report. Participants were askedto indicate experience with computers ranging from first timeuser to more than 6 months of experience. For those never usinga computer, the mean time to author their first lesson wasthirty-five minutes. For users with more than six monthsexperience, the mean time was 12.5 minutes. However, this dif-ference between the two groups was no longer observed afterseveral consecutive uses of the system.

Teachers had two general recommendations for improving thesystem. First, teachers expressed a need to create their owngraphics to correlate with the classroom program. They felt thiscapability would provide more compatibility between computer-based and regular instruction. Second, the system provides noexplicit session-by-session data on student gains. The solutionappears to be to collect information on selected variables andpair this with summary data in a manner that is useful toteachers. This data would be provided on a number of parametersand offer decision making points for the teacher.

28

Single subject designs could be used to examine the effectsof the authoring system. Response time, percent of trials cor-rect and trials to criterion would be dependent measures.Instruction, content, feedback, and reinforcement selectionswould be potential independent variables.

Teacher performance could also be monitored to determinewhether they were responsive to available data and changed thelessons as needed. The authoring system is easy for teachers touse. There is also a need to develop a data system that would beas easy for teachers to use.

Paul Hazan of the Applied Physics Lab at Johns Hopkinsindicated that one of the keys to success in this project was theuse of an interdisciplinary team. The authoring system was to bea tool that could stimulate the creativity of the users. Theteam started out by assessing current approaches to authoring andidentifying the strengths and weaknesses of existing systems. Asa result of the assessment of existing systems, one key featurewas to design a system that was easy to use. The needs oflearners were established and the technical options evaluated.

Several features were considered in designing the system.The ability to add graphics and to select and vary reinforcerswas considered essential. While authoring, the user can recallkey information from previous screens and preview a single lessonor a sequence of lessons. A maximum of ten lessons can be heldin memory for preview and editing.

The ability to design adaptive feedback to student responseis another feature of the system. Other considerations builtinto the system are auto responses, help features, and editingcapabilities. In creating a lesson, the author can visualize andedit content as well as create material from scratch. The systemprovides a step-by-step approach for ease of learning. Once thesystem is learned the user can author at a faster rate. Thesystem was designed for the 64K Apple. In that way schools canmake maximum use of the authoring capabilities with minimumhardware configuration.

In summary the authoring system as a vehicle for research aswell as an instructional tool was emphasized. The use of theauthoring system to examine process and content variables onlearning rates and patterns was illustrated.

ZUCKERMAN

Robert Zuckerman of Kent State University, discussed thedesign of the SPE.ED. Authoring System and the alpha and betatest procedures employed as an integral part of the softwaredevelopment.

The purpose of this project was to develop an authoringsystem that would allow teachers with little or no programmingexperience to design courseware for computer-assisted instructionof handicapped students.

The design of the authoring system was to take advantage ofteacher expertise allowing them to develop courseware efficientlyfor use by handicapped students. One major feature incorporatedin the design of the system was the capability to modify the run-time environment of courseware for students of differing abili-ties/handicapping conditions. By design, to utilize theauthoring system in the most effective manner, the teacher needsto identify 1) what content the students are to know, 2) how thestudents are to know the content and finally, 3) howinstructional content is to be presented. The SPE.ED. AuthoringSystem is a tool to assist teachers to mediate interactions withtheir students. The authoring system facilitates the developmentof the presentation of instructional content for purposes ofintroduction, practice, or testing.

The goal in developing the authoring system was not to haveall teachers using the computer in their classrooms but to pro-vide a tool that would allow teachers to design computer-basedinstruction without having to learn a high level programminglanguage. The SPE.ED. Authoring System was designed to allowteachers to develop courseware which could provide a match be-tween the process skills of the student and the instructionalcontent of the curriculum. The ability to develop coursewarethat is appropriate for the process skills of the student is toavoid common sources of why software fails. Inappropriate con-tent in available courseware and courseware content that thestudent has already mastered reflect the lack of appropriateinstuctional match between student process skills and the in-structional content of the courseware.

Authoring is viewed as a two stage process. In the firststage the teacher is responsible for selecting the appropriatecontent and identifying how the content will be delivered- Inthe second stage the authoring system is employed the vehicle fororganizing and entering the content presentation/interaction intoa form that is "played back" by the student. The authoringsystem must provide as much flexibility as possible to createdifferent kinds of lessons.

There are some trade offs in developing a user friendlyauthoring system. For instance, the developer has to decide how

much prompting or error recovery to provide. Prompting and errorrecovery are costly in terms of the space required for the codeversus the amount of space required for the program itself. Inaddition, the speed of a program will be greatly effected by theamount and manner in which prompting and error recovery areimplemented.

The concept of "frame" is utilized in the SPE.ED. AuthoringSystem to develop courseware. The frame is a unit of instructional interaction consisting of frame type identification,text/image to be displayed on the screen to the student, questiontype for query frames, correct response/answer for query frames,and exit path targets. A variety of different frame types areavailable to vary the instruction interaction available. Manydifferent strategies for instructional interaction by carefulselection and sequencing frame types. The frame type concept is"menu driven", different options are available for differenttasks in the instructional process.

The SPE.ED. Authoring System was designed to avoid a numberof pedagogical errors that often find their way into coursewareproduct. In authoring systems that do have pretest features manydo not force the teacher to avoid the violation of "pretestassumptions." For example: during a pretest new content shouldnot be introduced nor should practice or testing take place. Apretest is used to identify what the student knows at the point

IIIin time the pretest is taken. Many authoring systems do not havepretest features.

In the SPE.ED. Authoring System the user has the option toincorporate a pretest. When selected, the results of a pretestcan be used to branch the student anywhere in the lesson. Thecourseware author is responsible for determining branching patterns and targets followed based on the evaluation student'sresponses on the pretest. Pretest questions do not allow branching based on student responses. While constructing a pretestframe selection is limited to frame types pedagogically appropriate under a pretest condition.

When the courseware author indicates they are finished constructing the pretest, they then have access to a completevariety of instructional frame types. The manner in which theSPE.ED. Authoring System limits the design choices during theauthoring process helps the user to structure the presentation ofmaterial.

The software development phase relied upon tests of realityconstraints. The developers tried not to impose limits too earlyin the design process. A metafile structure was developed forthe authoring system as well as a uniform approach to the conceptof globals. Through the globals employed in the SPE.ED.Authoring System many features can be added enhancing the courseware functions served. The future integration and use of gra

phics is accessed through a global entered during the authoringprocess. The metafile structure allows transportability.Courseware files developed under one operating system can betransferred to run under another by transfering the file to thedisk format employed by the OS.

A two stage beta test involving specialists and teacherswill be conducted. Specialist can provide information about thespeed and efficiency of the system. Users will keep a log ofanecdotal data over 15 sessions and then complete an evaluationof the system rating the usefulness in developing computer basedinstruction, the features used, and other evaluation data. Thelogs will provide information for revising the software and willalso confirm the system design.

The alpha testing of the SPE.ED. Authoring System was aformative evaluation procedure and was viewed as an integral partof the software development process. The overall design of thesoftware was conceptualized, an IPO analysis and a "three-levelanalysis" that served as the software design reference weredeveloped. As routines were coded, alpha testing was used todetect logic and coding errors. As routines were integrated,alpha tests revealed any descrepancies in the conceptual designand the actual performance of routines. The alpha testing indi-cated that the access routines and other elements of the softwarewere consistent with the design concept.

Questions of product impact on handicapped students andtheir teachers will be undertaken after the product is dissemi-nated in final form. In the Beta Test, information about theperformance of the SPE.ED. Authoring System will be obtained.The Beta Test will be conducted in two waves. In the first wave,subjects will be "special interest, special skilled users" whocan provide commentary and feedback on the technical aspects ofthe authoring system. Subjects in the second wave will representthe target audience for the authoring system. Characteristics ofthese users include rudimentary knowledge of computers, and mini-mal programmiLly skills. One of the key factors and issues inconducting beta tests is to maintain a distance from the indivi-duals who are using the software and hardware products developed.Another issue that arises is when should investigators contactbeta test participants in order to elicit the response?

The authoring system will be in the public domain andtechnical assistance will be available to users.

e

SESSION 4: MAJOR TECHNOLOGY-BASED RESEARCH INITIATIVES

The Educational Technology Center (ETC) is the first of 17laboratories and centers funded by the National Institutes

of Education (NIE) to focus on technology education.The Center's task over a five-year period is to find ways of

using computers and other information technologies toteach science, mathematics and computing more effectively.

Dr. Jackson discussed the Center's research agenda andcurrent research investigations.

Gregory Jackson, Educational Technology Center, Harvard University

33

37

JACKSON

Dr. Gregory Jackson, of the Educational Technologies Centerat Harvard University, spoke about three research projectscurrently being conducted by the Center: Heat/Temperature,Videodisc, and Word Processing. He concluded with some back-ground information on the Center, and the underlying philosophyof technology-based research.

The premise of the Heat/Temperature study is that part ofthe reason students don't understand key concepts in science is

due to a gap between observation of what's happening and theirability to see the relationship. The question raised was: Cantechnology be used to help bridge this gap? In the case ofHeat/Temperature, heat is an extensive quantity and temperature,an intensive quantity, and the distinction between the two isobscure. Students tend to get bogged down with the experimentalprocess and they miss the cause/effect relationship. If studentscould have immediate access to what the relationship is, theymight better understand the concept.

The Center decided to investigate the problem because it metfour predetermined criteria: (1) teachers felt it was a hardconcept to teach; (2) students felt it was hard to learn; (3) ifthe concept isn't understood, it's hard to progress in science;and (4) it led to an interesting idea of how to use technology tomake such things easier to teach/learn.

A device was developed that would use a probe to measure thetemperature of water and display the values on an Apple screen,either singly or on a graph against time. Immersion heaters wereused to add measured amounts of heat to water. Teaching unitswith and without the equipment were developed, and pilot testswere run.

Teams sent out to collect evaluative data heard much nega-tive feedback regarding how the experimental lessons ran. Butpre-test/post-test results indicated that students working withthe equipment gained twice as much understanding of the concept.The researchers learned the following: (1) there are many detailsinvolved in bringing technology motivated by curricular interestto the point where you can study how well it pays off; (2) youmust be very careful how you measure success, as it is not neces-sarily equated with hapiness; and (3) even if an experimentdidn't go well, the technology may have helped.

The second project discussed was the use of interactivevideodisc to help students in developing hypotheses rather thanjust testing them. Eighty percent of users of interactive video-disc systems today are training operations in industry and themilitary. Existing programs are generally of two types: menusof different episodes of traditional CAI programs that branchbased on student performance. If interactive videodiscs come toschools at all, they are not likely to be much different from

34 38

110 training programs currently in existence.

Working in conjunction with the Children's TV Workshop,episodes from "3-2-1 Contact" and "Nova" are being adapted foruse on Interactive videodisc. The disc is currently in produc-tion and includes five episodes. It will be tested in six juniorhigh science classrooms next fall.

The third project, the Word Problems Project, was developedby the co-director of the Center. He was interested in thestructural differences and methods of teaching word problems, andin the differences students and adults have with word problems.Discussions with teachers proved that their perceptions of thedifficulties were quite different from his. For over a year,these issues have been discussed without any mention of tech-nology. The main thing learned so far is that often in order tocome to a productive use of technology you need to understand theissues of the subject matter being taught. If they are notclear, you can't expect technology to clarify them and solve theproblems. Discussions have moved toward how to get computers tomove from simple to complex concepts, to expose students to alltype of word problems, and how to use computer graphics to buildimages that explain concepts such as miles per hour and density.No experimental work has been done yet on this project.

The three research projects were selected as representativeexamples of what the Center is doing. The interactive videodiscproject is in the area of r.ew technology -- what is likely to beimportant over the next few years but isn't important now. TheCenter also looks at existing technology in the teaching of math(Word Problems), science (Heat/Temperature) and computing.

The Center came about in response to a request for proposal(RFP) from NIE. The RFP called for a Technology Center to solvenational problems in math, science, and computing, and serve theNew England region. It was to be a collaborative effort withmost research being done in the schools. Technology was recog-nized as being important, but had not yet become a focal point atHarvard. The Center proposal became a means of focusing atten-tion on research in educational technology.

One of the first governing principles in the Center'sresearch is that technology rarely raises new issues, but ratherforces us to look at older, underlying issues presented in a newguise. A second principle is that technology should beapproached from the subject matter. Instead of asking what amachine can do and then finding a place for it, researchersshould find out what isn't going well educationally and determinehow the machine can help. The third principle is that the Centerand all of its research projects, should be collaborative ef-forts. The Center comprises some twelve organizations liked bysubcontracts. Hopefully, these working relationships will con-tinue beyond the 5-year scope of the project. The researchgroups have generally agreed on issues, underlying theories,appropriate research questions, and ways to proceed.

-35 39

The Educational Technologies Center currently operates 12-14research projects. They run conferences, training institutes andseminars, and publish two newsletters and periodic reports.Their primary charge is threefold: First, to determine underwhat cirtl.umstances it makes sense to use technology in generaleducation; second, when it is appropriate, to decide what elsehas to happen; and third, to raise the level of discourse andcause people to think about technology in a different way and tointegrate it into bigger issues. The lasting effect of tech-nology will be through deep uses rather than superficial useswhich motivate students temporarily, and then pass away.

e

SESSION 5: THE USE OF ADVANCED TECHNOLOGIES IN SPECIAL EDUCATION

A presentation by COSMOS Corporation on a fifteen-month,SEP-funded study to identify current research in

robotics, artificial intelligence, and computer simulationand to explore the potential applications of these

advanced technologies in special education.

Gwendolyn B. Moore and Elizabeth A. Lahm, COSMOS Corporation

341

MOORE AND LAHM

Gwendolyn Moore and Elizabeth Lahm, from the COSMOS Corpora-tion, discussed a 15-month study on advanced technologies. Theproject, funded by SEP, examined three technologies -- robotics,artificial intelligence and computer simulation.

The two project objectives were: (1) to identify how andwhere these technologies were being used in settings other thanin special education; and (2) to determine which of the technolo-gies might hold promise for special education students.

Liz Lahm explained three underlying methodologicalapproaches: technology-push, demand-pull, and hybrid. The"tech-push" method focuses on the technology. A product isdeveloped assuming there will be a market for it when it's ready.The main problem is that such products may not directly addressthe needs of consumers. Conversely, the "demand-pull" methodbegins with the consumers needs and products are developed tohelp meet those needs.

The "hybrid" method, developed especially for the COSMOSstudy, combines parts of the other two methods. Existing appli-cations of the technologies are identified in fields other thanthe target field. Then forecasts are made regarding future usesof the technology in the target field. The "hybrid" approachfocuses on the application rather than the consumer or the tech-nology itself.

The steps in the hybrid approach include defining the tech-nology and locating its uses, documenting uses in an alternatefield, developing descriptions of similar uses in special educa-tion, rating those descriptions, and sharing findings of thestudy with potential users and developers.

The three technologies selected for study were robotics(which simulates human motion), artificial intelligence (whichmimics human reasoning and perception), and computer simulation(which mimics human sensory experiences). Applications rangefrom industry to home and school settings. Some 55 state-of-the-art applications were identified through the literature andthrough interviews with knowledgeable experts. Subsequent exami-nation led to the selection of 17 cases for further study. Addi-tional information was collected through site visits and exten-sive telephone interviews.

Gwen Moore discussed the research questions, the findings ofthe panelists, and the preliminary conclusions of the study.Each of the 17 cases was presented to a panel of ten specialeducation technologists. The following questions were answeredby the study:

1. What special education populations will benefit by eachtype of technology?

38 42

2. In what activities will each type of technology assistspecial education populations?

3. How many years will it take before each type of technology will benefit JO, 50, and 90 percent of thespecial education population?

4. What factors inhibit the implementation of each type oftechnology?

5. What dollar investment will be necessary to implementeach type of technology for special education populations?

Tables, showing data responding to these questions, wereshared with the audience.

Based on the panelists ratings and other materials examinedduring the study, the following conclusions were presented:

1. Robotics was thought to be most beneficial to thephysically impaired while simulation would be of morebenefit to the mild and moderate mentally handicappedand learning disabled. Responses for AI were morediverse, with two applications benefitting multiplepopulations.

2. The activity most frequently accomplished throughrobotics would be manipulating objects. AI would bestaid one's understanding. Simulation would be mosthelpful in experiencing things without the inherentdangers.

3. The panel felt that computer simulation would be thetechnology most likely to reach special populationsfirst. The 5 scenarios in that area were all projectedto reach 90% of the population in 10 to 20 years. Itwas projected that robotics would reach 90% of endusers in 15 to 19 years. AI again had the most diverseresponses, with one application reaching 90% of endusers within 15 years and other applications reachingonly 10% after 15 years.

4. The biggest barrier to implementation of both roboticsand simulation was expected to be cost. In the area ofAI, training and technical difficulties were expectedto pose the most problems. This.was because the scenarios would involve complex systems for which there areno relevant models in current day practice.

5. The technology expected to cost the most to implementwas AI, partly due to the large databases needed tosupport such systems. Four of the 5 scenarios wererated at over $500,000. Three of the 5 computer simu

lation scenarios fell between $200,000 and $400,000,with the remaining 2 rated at over $500;000. Only oneof the robotics scenarios was ranked in the highestcategory. The other two were rated between $100,000and $300,000.

The final report on the study is still being revised, so theconclusions presented are to be considered tentative. First,there is a great diversity of applications and uses of all threetechnologies, so generalizations about the technologies must bemade with care. Second: there are many lessons to be learnedfrom the current applications of the technologies in otherfields.- Several of these applications are technically feasibletoday and pose little or no problem for adaptation to specialeducation. Other applications will face significant cost, technical, or substantive barriers. Many of the barriers need to beaddressed through uses of the technology in other settings, notjust in special education. It is unrealistic to think that thespecial education community will be able to take the lead insurmounting the obstacles. This doesn't make the use of thesetechnologies in special education impossible, but it does extendthe timeline.

We must be careful not to let the technology overshadow theneeds of the students. In cases where a technology could beused, we need to ask if it is indeed an improvement over thealternative ways of addressing the same need. And regardless ofthe setting, the technology should be meeting a real need as wellas enhancing the education of the student.

SESSION 6: REPORTS OF RESEARCH IN PROGRESS

Brief reports of current research investigations by eightof the symposium participants. The researchers discussed

the purpose of their study, the design and methodology, and thefindings of their investigation.

Susan Jo Russell, Technical Education Research CentersRaymond G. Romanczyk, SUNY-Binghamton

Rick Brinker, Institute for the Study of Developmental DisabilitiesRichard D. Howell, The Ohio State University

Ron Thorkildsen, Utah State UniversityPhilip M. Prinz, The Pennsylvania State University

Roger Awad-Edwards, Palo Alto VA Medical CenterAl Cavalier, Association for Retarded Citizens

RUSSELL

Susan Jo Russell of the Technical Education Research Center,discussed findings from the survey component of a project calledMicrocomputers in Special Education: Beyond Drill and Practice.The project focuses on non-drill and practice uses of the micro-computers with students in grades K-8 who are classified ashaving either learning and/or emotional problems.

The term learner-centered software, used for non-drill andpractice applications, was defined as uses of the computer where:(1) students make choices about the goals of the activity orabout the strategies used to reach the goal; and (2) the feedbackthe user receives from the computer is informational rather thanjudgemental.

In planning the project there was a need to put activitiesin context. More information about the state-of-the-art in theuse of learner-centered software with this population was needed.The predominance of drill and practice software was welldocumented but information about the characteristics of peoplewho were using software other than drill and practice and thebarriers to implementing drill and practice were not welldocumented. The question guiJing the research was - Why isn'tthere more tool use of computers?

The survey had two components. The first was a randomsample of 50 school districts across the country selected fromall districts using computers for instructional purposes. Thesecond part was a selective survey of individual teachers or sys-tems that have promising uses of learner-centered software.

Russell reported on the results of the first part of thesurvey. The topics covered were the types of applications,strategies for helping teachers learn about computers, barriersencountered, factors facilitating use, and outcomes or benefitsfor students. Telephone interviews were conducted with specialeducation administrators in each school district. A follow-upinterview was conducted with a teacher in the district nominatedby the administrator as being experienced and knowledgeable aboutthis kind of computer use.

The administrators, in 83% of school districts reporteddrill and practice as the primary use. About one-fourth of theadministrators reported using word processing software with stu-dents. This is the only tool software mentioned. There was nomention of databases or spreadsheets. Few administrators couldgive details on how this software was being used or why. Almostall administrators said the school system had provided some formof training, usually designed for regular education teachers.

Teachers were identified by administrators as barriers andas facilitators. One-fifth of the administrators talked aboutstaff resistance as a problem while more than one-fourthindicated that staff enthusiasm was one of the ways things weregetting started. Administrators and teachers both mentioned lackof good software as a primary problem.

The most interesting thing about administrators was whatthey said about benefits. In general, administrators had verylittle idea of what the possible benefits were. Stated goalswere not found in any school district. A large number of theresponses were of the type: "The reason we're using computerswith our students is that it's the wave of the future, we can'tbe left behind." The other benefit most often cited was motiva-tion. Forty percent said computers were used because they mightmotivate students. Very few administrators mentioned any sort oflearning outcomes.

The teacher interviews confirmed and also expanded thesefindings. Because something was classified as non-drill-and-practice, did not mean that it was being used in a non-drill-and-practice way. There were instances where word processing wasused to copy spelling words or to write sentences for spellingand not in any other way. Only one-quarter of the teachers usedword processing as a writing tool. Problem solving software,such as LOGO, was often used as a reward when other work isfinished. It was not viewed as a learning tool.

The second thing that became clear with the teacher inter-views was that there is a very low knowledge base about what ispossible, what computers can be used for. People simply didn'tknow what spreadsheets or databases were. The third finding thatbecame clear was the inappropriate nature of the training mostteachers had received. A large number had received training inprogramming and BASIC. None were using programming with theirstudents. The few who had learned in an informal setting from apeer (teacher, student teacher, family member) expressed mostsatisfaction with their training. Teachers did not describe thelack of training as a barrier. Only 6% indicated lack oftraining as a barrier. Lack of resources was a very majorbarrier mentioned by one-third of the teachers.

The survey raised several questions that require furtherresearch. For example, what distinguishes people who uselearner-centered software from those who use only drill andpractice? Most teachers started with drill and practice. So thequestion is for those who went beyond; what were the factors. Itseems to relate more to the literature about successful managersand entrepreneurs than to the educational literature. A secondquestion is what alternative training models might be effective.

Finally, how do we cope with the lack of curricular fit.There is a pervasive sense that the way people are using compu-ters is outside the curriculum.

ROMANCZYK

Raymond Romanczyk of the State University of New York atBinghamton reported on two studies conducted at the Children'sUnit for Treatment and Evaluation. The first study involvedseventeen children diagnosed as schizophrenic, autistic, andseverely emotionally disturbed in a visual discrimination task.An errorless learning paradigm that had a teacher and a computercondition was used to present the task. Researchers tried tocontrol for every variable other than the presence of the person.For instance, if a touch sensitive screen was used in the compu-ter condition, a stimulus card behind a sheet of plexiglass wasused in the teacher condition to duplicate the tactile stimula-tion. Each child was run through both conditions to counterbalance and control for sequence effects. A very high qualityvoice synthesizer was used and the teacher was provided timinginstructions as to the pacing of the session.

Basically, no differences were found between the effective-ness of the teacher and the computer for this particular popula-tion. That finding was in contrast to previously publishedresearch. However, there were individual differences. Somestudents did better with the teacher or the computer. Behaviorslike self-stimulation, acting out, and maladaptive behavior weresignificantly less under the computer condition.

Classroom measurements of predictor variables were taken todetermine who would benefit. The findings indicate that thestrongest predictor was a simple one. Classroom observation ofchild noncompliance to task direction indicates that the higherthe noncompliance, the better the student does with the computer.In general noncompliance appears to be very strong. This mayindicate escape motivated behavior and the interaction oflearning history. One related finding in the literature is thatfor autistic individuals, self-stimulation appears incompatiblewith learning. That is, the more self-stimulation being engagedin, the less likely the child is to acquire the information beingpresented.

The correlation between errors and self-stimulation with theteacher was .77. With the computer the correlation was .04. Sothis characteristic associated with autistic individuals is pre-sent only under certain conditions. With all diagnostic groupsthere were signifcant correlations between patterns of self-stimulation and errors with the teacher. No group showed signi-ficant correlations with the computer for the same two variables.This suggests the need to rethink some of the supposed relation-ships between behavior, performance and learning. Social learn-ing history may be confounded with the specifics of the learningtask.

The second study involved generalization. Using simplemathematics tasks, researchers asked the question "what happenswhen we use a more functional task?" The paradigm was to pretest

44

children diagnosed as severely emotionally impaired. The teacherthen presented flash cards and worksheets followed by computerinstruction. It was a three level program. At Level 1 problemswere presented with counting prompts. At Level 2 countingprompts were presented only after errors. At Level 3, nocounting prompts were presented. There was great variability innumber of sessions conducted and no significant correlation withperformance. For four out of five subjects, the generalizationto flash cards was good. However, for the fifth subject pre- andposttest scores were almost identical.

Generalization to worksheets was also good and performancemuch improved from baseline except for fifth subject. Thatstudent exhibited the same pattern of generalization but stillhad high performance on the computer. This was a child who wasvery fearful of any contact or feedback with adults. She hadacquired skills on the computer but there was a whole learninghistory that was preventing generalization from taking place.This example underscores the potential of computers as diagnostictools.

Another study involved learning disabled students in mathinstruction taught by teacher and by computer. Concurrentpsychophysiological measures were taken in both conditions.There was no difference in measures of rate and accuracy betweenteacher and computer until a simple token reward condition wasadded. The token condition raised performance. When the rewardsystem was removed, the performance of the computer group fellapart. However, performance with the person maintained. Thisfinding implies that computer variables as manipulations may bevery fragile and caution is addressed in applied settings.

Given the software used and the findings, the researchersare proposing some standards as minimum for data collection inCAI programs. Archival data of each session such as, date,session time and duration are collected. Distinctions betweenpresentation time of stimuli versus time spent in responding tostimuli versus consequence time are made. Parameter controlssuch as the duration of presentation, number of sessions, problemboundaries, feedback, voice presentation, reinforcement type andschedule, and presence of prompts are routinely collected. Datacollection on these parameter control permits generalization offindings.

BRINKER

Rick Brinker discussed a contingency intervention projectconducted at Educational Testing Service. The purpose of theproject was to make handicapped infants explore their world.Infants in the study were organically impaired. That is, therewas an organic condition associated with their handicap. The twofactors associated with the organic conditions were a reductionin movement patterns which made it more difficult to contactobjects in the world and a social system problem. Parents hadnot received positive information about their child's capabili-ties. The microcomputer provided the capability for a quick turnaround of information and helped in handling the expectencies ofthe social problem. Parents could see a graphic portrayal ofbehavior following the intervention.

The microcomputer was viewed as a prosthetic device. Theinfants studied did not hold an object in their hands or bringthe object to their face, and they had no visually directedreach. An understanding of objects is a very important basis forsymbolic development. That understanding comes about throughcontact with and manipulation of objects. In the intervention,investigators prosthetically arrange an environment that wouldmake consequences and objects perceptable to children. In thisenvironment the child could see that a consequence was directlyrelated to their own activity. The purpose of the prostheticenvironment was to encourage the infant to explore. If a childis in a situation in which he controls interesting consequences,the child learns to explore and to control consequences.

In the literature on learning it is difficult to find amodel for analyzing that level of data with an individual sub-ject. Most of the infant learning literature is of group com-parisons where control and experimental groups receive someseries of similar stimulations. It's difficult to work at thelevel of single subject design and to look for learning on adaily basis and design the next intervention. The experimentaldesign used first established a baseline where infants activity(pulling, hiding, vocalizing) was counted. Then one of theactivities is made to produce consequences. Later contingenciesare reversed.

The developmental hierarchy is described by Piaget as thedevelopment from primary circular reaction to means/ends behaviorand is usually completed by eight months of age. First, theinfant learns that they can make things happen. The next step isa secondary circular reaction in which the infant controls in aspecific way. The final level is the sequencing and subsumingbehaviors as tools to explore other consequences. Once an infanthas visually directed reach and is exploring the world of objectsthe infant no longer needs the reinforcement environment.

The microcomputer also provides access to an archival data-base. Contingent probabilities are used to see how frequently aninfant at different phases of learning emits an arm and then aleg response in a given contingency. Alternating responses wouldmake sense in early sessions while subject was still exploring.The frequency of alternations decreases over time, and there isan immediate increase in alternations when the contingency ischanged. That increase indicates exploration.

Ten moderately handicapped subjects received between two andeight months of intervention. Each made at least a one monthgain in mental age per month in the intervention situation.There are some parallels between these findings and findings withschool-aged children. An example is the potentially negativeeffects of reinforcing students for a high error rate or rein-forcing impulsively. The capability to maintain computer-baseddata provides a clearer performance base on handicapped students,and gives a much clearer description of performances than everbefore.

e HOWELL

Richard Howell of The Ohio State University discussed aseries of investigations focusing on CAI use with mildly handi-capped, learning disabled students in the area of mathematics.He indicated that the question is not so much whether CAI works,but where drill and practice belongs in the educational process.Almost 60% of programs used in special education are drill andpractice based. CAI software can be modified in terms of rate,the difficulty of problems presented, and in terms of the opera-tions. There are several areas that require attention. First,educational games have a number of variables, a number of stimu-lus items (graphics, sound, display, characteristics) that shouldbe studied. Second, the focus of most software is not onaltering the strategies by which the learner approaches the task,but rather on the manipulation of content related items by thelearner.

In the pilot study, a single-subject, ABAB type design wasused to determine if a common form of drill and practice softwarewas an effective intervention for the acquisition of multiplica-tion facts by learning disabled students. The drill and practicesoftware used did not featurs any strategy-based learning. Datagathering involved recording the number of errors per ten prob-lems and the amount of time on task during each observationalsession. The data points represent an average of both errors andtime over sessions. The baseline recording was done by a randomsample of multiplication problems presented to the student usingpencil and paper. The intervention was done by observing stu-dents performance as they answered ten randomly generated prob-lems. The latency between presentations and problems was heldconstant.

The independent variable was an experimental procedureintroduced in the first intervention period. Researchers lookedat strategies students were using to solve problems. In solvingmultiplication problems, children were using a base five refer-ence to approximate a number. The use of a base five referentpoint effected latency and also increased the possibility forerror. The intervention involved thirteen sessions. The base-line period lasted 3 sessions and was followed by 3 days ofintervention. A return to baseline condition followed. Thefinal intervention period lasted 4 sessions.

The average number of errors per session increased from 1 inthe first baseline session to 4 in the second baseline session.During the first intervention the drill and practice software wasintroduced. The subjects used the software for twenty minutesand errors decreased to 1 by session 6. After the baselinecondition was reintroduced the error rate increased.

The results indicate that drill and practice software mayhave an initial but transitory effect on the number of errors andthe amount of time required to successfully complete multiplica-

tion problems. It appears that without a specific interventiontreatment that changes the strategy with which the studentapproaches problems, any gains made during the computer inter-action will not hold over time.

The directions for future research seem to indicate the needto use CAI software that introduces a strategy for solving multi-plication problems within a tutorial-based framework. The soft-ware may have a drill and practice compenent as a reinforcementstrategy. Subsequent studies will introduce a new compensatorystrategy using CAI in tutorial-based software. The child willwork to a plateau, maintain that level, and then move to the nextlevel with the compensatory software.

THORKILDSEN

Ron Thorkildsen described the Social Skills Training Programdeveloped at Utah State University. The Social Skills TrainingProgram teaches cooperative interaction skills. Skills includegetting involved, being involved, and being positive. Strategiesfor remaining calm in negative situations are also taught.Effective communication attributes such as use of correct into-nation, correct body language and appropriate vocabulary areintegrated into the skill areas.

The Social Skills Training Program used the videodisc topresent examples of appropriate and inappropriate behaviors inthe target skill areas. In addition to the videodisc there wasprint material with the program. At the outset of the projectthe intent was to present all information in a videodisc format.After five field tests with interactive videotape, it was decidedthat print material was preferable to a computer or videodiscscreen when information presented was to be read.

The program is designed for small group instruction in upperelementary resource rooms. Materials are based on a directinstructional model. In the field test phase, teachers workedwith five students. The teachers used the videodisc to presentexamples of social skills. After each example, there was a dis-cussion of the scene presented. The videodisc was also used forimitation and structured role-playing.

A behavior management system is an integral part of theprogram. The four month field test involved pretesting, imple-mentation of the program, and posttesting. Six classrooms wererandomly assigned to the treatment or control group. There wereEifteen students in each condition. All schools had some kindof social skills behavioral management system. The control grouphad no additional social skills training.

Natura]istic observations and sociometric testing were usedthroughout the four month process. Observers looked at negative,positive, and alone behaviors. Resource room students spent mostof the day in regular classrooms. The sociometric testing wasconducted in the regular classroom. All children in the classmade a list of classmates and indicated who they like to playwith most and least. In sociometric nominations students wereasked to list the 3 people they liked most and least. Self-esteem testing, using a behavioral checklist, measured changes inspecific behaviors and criterion checklists measured skilllearning.

The sociometric testing revealed some differences. A com-parison showed statistical difference (three-fourths of a stan-dard deviation) between the experimental and the control group.Something was causing students in experimental group to be betteraccepted by their peers. There was no difference in self esteembetween groups and very little difference on the behavioral

checklist. However, there was a large statistical difference oncriterion checklist. The students did learn the skills.Teachers reported little difference in behavior in peer inter-actions.

In the naturalistic observations, the control group startedout a lot higher and more positive than the experimental group.Trend analyses showed the control group dropping and the experi-mental group going up significantly. The interaction between therepeated measures and the treatment was significant.

Thorkildsen predicted increased group use of computers andvideodiscs. It is a good use of a scarce resource that worksbetter in a lot of situations.

e

e

PRINZ

Philip Prinz of the Pennsylvania State University describedthe development of a computer/videodisc-based system for teachingwritten language, reading, writing, and general communicationskills. The basic philosophy underlying the development was thatreading and writing should be taught in an interactive mode andwritten language in a discourse or conversational mode. Themicrocomputer/videodisc was used as a tool at the interface ofwritten language instruction and general communication instruc-tion.

The project, called the ALPHA Interactive Language Programinvolves special education teachers, speech and language patho-logists, LD specialists in classrooms and resource rooms. Theprimary target group was hearing impaired children. The programstarted with a preschool population and expanded to includeelementary aged hearing impaired students. A pilot study wasalso initiated with multihandicapped children who had severecommunication (expressive output) problems.

A language/reading assessment battery was used for pre- andpost-testing measurements. A modes imitation test or sentenceimitation test sampling various modes - American sign language,finger spelling, simultaneous communication, spoken English andspeech only was administered. Once the primary mode of communi-cation was determined, all other subtests were administered inthat mode. There was a single word receptive and productievocabulary measure, a sentence imitation test, a grammatic com-prehension measure, and a generalized vocabulary reading test inthe battery. A teacher and parent language communication ques-tionniare and direct observations of communication interactionsin the classroom provided additional information.

In general, results indicated no gains in reading level butlarge gains in the use of computer hardware. In terms of writtenlanguage and general communication, children make significantgains not only in terms of reading vocabulary acquisition butalso specific gains in general language. Researchers alsostudied the interactions between the teacher, child and computer.The interactions were videotaped and the structural andfunctional aspects of the interactions were analyzed.

The current research objectives include determining theeffects of microcomputer instruction, the amount of instructiontime, and the effects on the development of written language. Asecond area of interest is the effect of the location of thecomputer in relation to achievement. The effects of classroomversus resource room placement are contrasted. Another area ofstudy is the impact of computer generated sign language onlearning. Children receive half of the test items with the signlanguage graphic and half exluding the graphic.

In the videodisc study the researchers will contrast com-

e

e

puter generated graphics and video excerpts. Classroom teacherswill be involved in inservice training and the effects onlanguage and written communication will be measured.

AWAD-EDWARDS

Roger Awad-Edwards described the work of a multidisciplinaryteam at the Rehabilitation Research and Development Center of thePalo Alto VA Medical Center. The purpose of the project was toestablish the concept of using an industrial robot, enhanced withvoice recognition and voice synthesis capabilities, to restorelost upper body functions to individuals. The goal was indepen-dence for both the physically disabled patient and the caregiverby augmenting, supplementing or replacing lost function. Theteam seeks to demonstrate the feasibility of this concept techni-cally and as a methodology for examining the relationship betweensmart machines and people.

The system consists of off-the-shelf hardware and customizedsoftware that is frequently adapted. The development of a human/robot team drew from many disciplines -- psychology, criticalpath analysis, education, and human factors. The first step wasto establish the parameters involved in a voice controlled robo-tic aid termed interactive robotics. Safety is a critical issue.There are several ways of addressing safety, including softwaremodifications, redundancy of safety stops, and emergency stopprocedures.

The methodology developed is termed interactive evaluation.It consists of four prototypes that assess technical and clinicalparameters simultaneously. The assessment is done in a cyclicalfashion. Milestones provide direction and guidelines for futureresearch, information, products, software, and principles thatcan be transferred to subsequent development. The superstructureincludes the analysis of robot factors and human factors fororganizing the research.

In order to generate a humane/machine profile, researchersassess the machines performance and it's impact on human charac-teristics. In turn certain human performances dictate require-ments for current and projected technology. Potential users,quadraplegics, are involved in the development and design phases.for our analysis included robot factors, human factors fororganizing our research.

Three stages in the development of the interactive evalua-tion prototype were: feasibilty, utility and potential market-ability, and diffusion of innovations. A major goal of theproject was to assess user acceptance and how it can be facili-tated. The users identified four major categories of tasks forwhich they would like a robotic aid. One of the tasks was a mealserving task. They also wanted the robot aid to help in voca-tional tasks such as opening file drawers and retrieving files.Recreation was another category. For example, real time voice-control and preprogrammed motions can be used for painting.

The approach used was an androgogical, not a pedagogicalone. Ninety-five percent of the work was done with adults; indi-

e

o

viduals whose sense of self direction is a key factor in using anaid. They want immediate not delayed application of learning.Five training sessions, 60-90 minutes in length, orient users tothe robotic aids. At the end of every session there is a stan-dardized proficiency task that provides tangible outcomes. Forexample, the first standardized task is to give yourself a drinkof water. There is also a program of peer training. The presentvoice operated system has 50 commands and is speaker dependent.The key to success with voice control is achieving consistencywith the voice.

The concept of robotic aids has been extended to includerobotic technology for health and human services. A three phasestudy using interactive and silent observation, is being con-ducted in a nursing facility. Robots are used primarily as atool. It was demonstrated that older people in nursing homescould use this technology. Other applications include trans-porting, lifting patients, ambulation, and vital signs moni-toring. An attitudinal assessment of health care professionalsand technological engineers was incorporated into the study.Researchers found positive attitudes if the professionals areproperly oriented to the technology.

In the future, a project for using this technology will beconducted to determine the extent of crossover from the currentpopulation to school-aged populations.

e CAVALIER

Al Cavalier, Director of the Bioengineering Program at theAssociation for Retarded Citizens indicated that national ARC hasa good field network that defines needs for researchers. Athrust in the last few years has been to increase independencethe most severely handicapped persons allowing them to partici-pate in community living. One of the limiting conditions fre-quently identified is a fundamental one of independent toileting.Incontinence is also a serious problem for individuals withhigher order skills employed in vocational programs. Individualsare often denied access to these programs because of lack ofcontrol for independent toileting.

The study of incontinence is focused on individuals who failto learn complete independence in toileting and do not haverelated medical or organic problems. Typically, training pro-grams usually involve increasing the level of fluids and therebyincreasing the opportunities for learning to take place. Thesubject is then taught a series of behavioral steps and throughphasing and physical guidance can achieve a criterion. Thestimulus is internal and focuses on a single point in thetoileting process.

Technology can be used to increase the speed with which a

toilet training program achieves success. The application is agood example of technology used temporarily to impart a skillthat a client can learn and then removing technology from thesituation. There is also a larger application to individauls whohave a permanent loss of ability to sense and control bladderfunction.

There is a theoretical base for this development effort.The toileting sequence involves stimulation in the bladder, theelimination response, and the consequences. Clients have toattend to three sets of stimuli and they are attending to onlytwo. The technology is designed to assist with the stimulus thatis lacking. By continuously monitoring the condition of thebladder, the technology senses the fullness of the bladder forthe person. It is a redundant cue.

The model has three components: a sensor on the body, a

black box or main logic unit, and a feedback device. The feed-back device returns an external signal to the client. The modeluses technology to externalize an internal stimuli for a shortperiod of time. That external stimulus is redundant with theinternal stimulus. The educational research in discriminationlearning indicates that a relevant redundant cue can be faded aslearning takes place. The logic unit has a signal that amplifiesthe signals coming from the transducer on the persons body. Thesignals are echoed back to the logic unit by body tissues todetermine whether an external stimulus should be provided. Thesensor must be placed on the body to have an unobstructed inter-nal view to the bladder.

The ARC researchers are working with untrasonic scientistsat NASA to map out bladder locations in varying body types and todetermine the variation between a full and a not-full bladder.The first phase of development is to define the parameters ofwhere the bladder is and what occurs when it is full. A proto-type using these measurements will be developed and alpha testsconducted. The logic unit will then be reduced to the size ofwalkman cassette recorder. The client will wear the unitattached to a belt.

With the powerful advances in sensor technology, eventuallythe sensors will be reduced to the size of a bandaid that can bepeeled off a strip. These advances will make devices far moreflexible and easy to use. The research program has selected twofield test sites, a vocational program and a public schoolspecial education setting.

Market research suggests a wide range of use for the device.It is estimated that in addition to 100,000 people with mentalretardation, there are five million persons that have bladderproblems who could use the aid. One area that has long plaguedbehavioral psychologists can now be assisted with the use ofmicroprocessor-based technology.

SESSION 7: FUNDING FOR TECHNOLOGY-BASED RESEARCH

A panel presentation on funding for technology-based researchwith representatives of government agencies, foundation

and industry funding resources. The panelists discussed currentprogram initiatives and alternate funding

strategies.

Mary K. Leonard, Council on FoundationsJ. Richard Taft, The Taft Group

Edward Esty, NIE/SBIRDavid Gray, NICHD/SBIR

Richard Johnson, NIHR/OSERSMartin J. Kaufman, Director, DES/SEP/OSERS

LEONARD

Mary Leonard, Director of the Precollegiate Education Pro-gram at the Council on Foundations, discussed trends in thefoundation giving and grant seeking strategies. The Council onFoundations is a membership organization of approximately 1000charitable foundations and corporate giving offices. Councilstaff does research on the field of philanthropy and providesprofessional development for staff and trustees of foundations.Other components are a strong lobbying program and an activepublic information program.

Leonard described some trends in foundation giving. Firstthere has been a slight expansion in the dollars available forfoundation giving. The dollars available for education haveincreased 8.5 percent from 1983 to 1984. There is also a largeexpansion in the number of grant seekers. Historically, founda-tions have funded in higher education institutions. Severalyears ago 96 percent of the money going to education went tohigher education. Foundations are increasingly more apt to giveto public institutions. There is also an increasing sophistica-tion in the development business. The hiring of professionalgrant seekers and development officers by institutions is morecommon. Foundations are trying to become more sophisticated aswell. The Council encourages members to become more focused intheir grant seeking. There are more cooperative grants or grantsshared by a number of foundations. Foundations are also lookingvery favorably at consortiums of organizations such as a localschool system and a local university. Challenge grants areanother increasingly used tool. These grants increase relianceon outside sources of funding in addition to the foundation orcorporate dollars. Foundations are targeting their money morecarefully. The field of philanthropy is trying to professiona-lize itself. There are about 24,000 private foundations. Onlyfifteen hundred are staffed and most at a level of two or threepersons. Foundations are interested in learning more about boththe subject matter of grantmaking programs and the process ofselecting, evaluating and providing technical assistance to gran-tees. We are also seeing more interest in the corporate sector.There is an increase in corporate matching gifts programs.Finally, we see an increase in the provision of noncash assis-tance. Corporations are moving into the area of program relatedinvestments, they will lend money for a development project.Other foundations are using thei_ auspices to convene meetings ofinterested parties. They are providing technical assistance fornew, grassroots organizations.

Leonard reviewed some of the basics of foundation funding.The most basic piece of information for grant seekers is to getto know The Foundation Center. The Center is a sister organiza-tion to The Council on Foundations. It is set up to provideinformation about foundation and corporate giving to the grantseeking public. The Center provides the information needed toapproach foundations successfully. Leonard advised grant seekers

59 63

to contact funders early on, before a project design has beendetermined. This gives foundations an opportunity to provideinput in the development phases. It is helpful to develop a

brief synopsis of your work that presents the information in a

precise, comprehensive way. It is also helpful to contact mul-tiple funders. Funders expect this. Finally, it is helpful to afunder to be informed about the progress of your project. Whenapproaching any funder, do your homework first. It is veryimpressive to an organization if you know as much as possibleabout them and what they do.

TAFT

Richard Taft, President of the Taft Group, provided a

Perspective about the philanthropic field in general and sometips about fund-raising. He indicated that only about fivepercent of all proposals submitted to foundations get funded.Funding at the federal level runs about eighteen percent. Onething that is important to grasp is the magnitude of the fundingindustry, the diversity of that industry and its complexities.There is 74 billion dollars a year in private philanthropy in

this country. Non-profit organizations, of which there are about800,000, constitute a large percentage of our gross nationalproduct. Taft pointed out that the competition in the fund-raising business has been intensive under the present administra-tion's emphasis on private sector activity.

Funding by foundations and corporations is only part of thebusiness. The bulk of the money comes from individuals. Approx-imately 62 billion dollars comes through bequests, contributionsto charity, and major gifts. Individuals are the big factor inthe philanthropic field.

There is a strong professional element to the foundation andcorporate field of philanthropy. Often successful fund seekingis based on contacts, who you know and the relationships youdevelop. Foundations break down into five areas. There are thebig generial research foundations like Rockefeller, Ford andCarnegie. They ususally have sizeable assets, holding 60-70percent of all assets in the foundation field.

There is anothe!: group called special purpose foundations.They have chosen to concentrate in one particular area such ashealth care or educa1tion. There are over 20,000 family founda-tions. Essentially' these foundations are extensions of oneindividual's philanthropy. Another kind of foundation is thecommunity foundation. It is generally a group of trusts at thecommunity level, administered by community leaders. Corporatefoundations are extensions of corporate giving. Two types offunding in corporAtions are direct giving and through founda-tions. Another way to develop funding is through endowments.Individuals and family foundations with a special interest in the

area often endow a research project. It offers a trememdouspotential for funding.

Grantsmanship is a process of market research. Proposalsare not the best route. In most cases you begin with a conceptnot a proposal, then contact the foundation, and develop a dia-logue. The one thing to remember is it's easier to get a hearingif you have contacts in the foundation. Foundations give topeople and to ideas. The way to demonstrate that you have a goodidea is to explain in a short letter or executive summary. Aproposal is in a sense a hinderance to the process of communica-tion. It cuts off the dialogue and discounts the foundation'sexpertise.

ESTY

Ed Esty, Director of the Small Business Innovative Research(SBIR), the Department of Education, spoke on funding availableunder that program. The Small Business Innovation DevelopmentAct was passed in 1982. Essentially the law says that anyfederal agency with an extramural research and development budgetof more than a hundred million dollars a year is required by theAct to set up a Small Business Innovative Research Program.There are now twelve such agencies; the Department of Educationis one of them. The percentage of money to be set aside changesfrom year to year. In the first year, FY83, .2 percent wasset aside. This year it is a full one percent. All participa-ting agencies set up a three phase program.

In Phase 1 people are given a small amount of money for sixmonths, to demonstrate or determine the feasibility of an idea.Proposers must be small profit making businesses of 500 or feweremployees. People who compete successfully for Phase 1 con-tracts are then eligible to compete for Phase 2 funding. Awardsare limited to two years and average about $200,000. Phase 2

involves conducting the full scale R&D that has been determinedin Phase 1 to be feasible. The third phase is not ordinarilyfunded by the federal government but by other sources of venturecapital. A document called The Pre-Solicitation Announcementprovides information on the program FE all twelve agencies. Itcomes out quarterly and lists the agencies having competitions,the release date, and the technical topics that the agency isinterested in. All solicitations from the twelve agencies arevery similar in format, making it easier for small businesses tobe involved in more than one competition.

GRAY

David Gray, of The Human Behavior and Learning Branch of theNational Institute of Child Health and Development, talked about

that agency's funding programs. NIH consists of eleven insti-tutes and has numerous mechanisms for funding. One is the SmallBusiness Innovative Research Grant and Contract Program. Eachyear there are three cycles of funding under this program. Therole of the program officer is to review the initial two or threepage synopsis of an idea and to set up a mechanism for trackingthe proposal.

The Human Behavior and Learning Branch is currently workingin the area of learning disabilities and encouraging people todevelop software for diagnosing and treating reading problems.Another area is the development of software or educationalmaterials for exceptionally bright children. A third area ofinterest is prevention of accidents and promotion of health.

JOHNSON

Richard Johnson, of the National Institute of HandicappedResearch in the Office of Special Education and RehabilitativeServices, discussed the institute's priorities. NIHR supportsresearch in all areas of disability, primarily in those ageranges that relate to rehabilitation. All NIHR grant announce-ments appear in the Federal Register. Johnson discussed severalpending announcements. One is for rehabilitation and trainingcenters to work in the areas of elderly disabled, improvement ofindependent living programs, rehabilitation of psychologicallydisabled persons, and community resources. Another currentannouncement is for a rehabilitation engineering center on tech-nology for blind and visually impaired persons. Three researchand demonstration programs on delivery of rehabilitation engi-neering services, computer adaptations for severely disabledpeople, and economics of disability will also be funded.

NIHR has innovative grant programs to test new concepts ormethods of working with the disabled and to purchase new devicesand evaluate them for use with different disabled populations.An annual field-initiated program funds research and demonstra-tion programs which have a direct bearing on the development ofmethods, procedures and devices for the disabled. A unique pro-gram at NIHR is a research fellowship program of grants to indi-viduals rather than institutions. The five priority areas inthis program are community retardation services, transition andemployment, early intervention, medical research and disabilitystatistics.

KAUFMAN

Martin Kaufman, Director of the Division of EducationalServices in the Office of Special Education Programs, gave an

overview of SEP directions for the coming year. He indicatedthat the field-initiated research program will be announcedearlier for next year. According to SEP's projected schedule,the notice for the training, demonstration and, research programsof the Office will be issued toward the end of July.

In addition, SEP will probably focus on two broad targetareas next year. One area is likely to involve collaborativework between researchers, school systems, publishers, and hard-ware providers on the issue of technology integration. Thepurpose is to find examples that integrate technology into theclassroom and demonstrate that integration is possible. In theshort run, examples of effective technology integration areneeded in the literature. A second area is likely to target onadvanced technologies for use in making augmentative devices orservices available to low incidence populations who currently arenot receiving the benefits of existing technology. The issue isof access to devices and the need to stimulate ways of makingdevices more available to handicapped individuals.

RESEARCH SUMMARIES

68

SUMMARIES ARE ORDERED ALPHABETICALLY

Philip Archer* Roger Awad-EdwardsMichael M. BehrmannRandy E. Bennett

* Richard P. BrinkerColleen A. Haney Bruno

(See Egolf)** Douglas Carnine

Phillip Cartwright* Al Cavalier

E. William Clymer** Richard Deni

(See Brinker)Donald B. Egolf

** K. G. Engelhart(See Awad-Edwards)

Gail Fitzgerald* Mark B. Friedman

Robert Gall** Bud and Dolores Hagen

Carol Hamlett* Ted HasselbringJacqueline Haynes

* Paul L. Hazan(See Panyan)

** Alan Hofmeister* Richard D. HowellJeffrey W. Hummel

* Kathleen M. HurleyJames Jurica

(See Howell)* James A. Kulik* Elizabeth A. Lahm

(See Behrmann, Moore)* Joseph P. Lamos

Charles MacArthurGail McGregorDavid MaloufBeth Mineo

(See Cavalier)** Judy Moll* Gwendolyn B. Moore

Catherine Cobb MoroccoJune E. Morris

64

63

Doris NaimanSusan Neuman

(See Morocco)Linda E. O'Donnell

* Marion Panyan* Philip M. Prinz* Herbert Rieth* Raymond G. Romanczyk

Teresa Rosegrant* Susan Jo RussellRegina H. SaponaGilbert Schiffman

(See Panyan)** Ken Seeley* Melvyn SemmelJack M. SpurlockRichard Steele

* Ron Thorkildsen** Greg Turner

Michael WolfeJohn Woodward

(See Carnine)* Robert A. Zuckerman

* Individuals who made presentations** Individuals who submitted Research Summary but who were unable

to attend.

70

65

ARCHER

STUDY Identification of Effective Implementation Strategiesfor Integrating Microcomputer Instruction into On-going Services for the Handicapped (ProjectINTERFACE)

SUMMARY "Project INTERFACE" is based on the premise thateffective communication in special education willcome only as a result of understanding the mosteffective implementation strategies for microcomputerusage and its capability to help special educationstudents. The purpose of this 2-year collaborativeproject between BOCES/Long Island University (LIU)and local education agencies (LEAs) is to determinethe most effective implementation strategies forintegrating microcomputers into ongoing educationalservices for the handicapped. The objectives addressspecific issues directly related to the project'spurpose.

Objective I - To determine those contextual andbehavioral variables required for effective integra-tion of microcomputer technology into ongoing educa-tional services for the handicapped.

Objective II - To develop and maintain mechanisms forthe particigating institute of higher education (Long

Inter-mediate

University at C.W. Post Campus) and the nter-mediate Educational Agency (Nassau County BOCES) tosynergistically integrate their research and trainingofferings.

Objective III - To develop and maintain demonstrationpracEiFl7Eites within a school of the Nassau CountyBoard of Cooperative Educational Services (BOCES) andlocal school districts which will incorporate thosecontextual and behavioral variables ascertained infulfilling Objective I.

Objective IV - To determine the impact of a col-laborative training effort on the performance ofteaching and administrative trainees in special edu-cation.

Objective V - To develop and implement mechanisms forthe two collaborative programs (Nassau BOCES and LongIsland University at C.W. Post Campus) to disseminateinformation on the project's research and trainingefforts and results.

More specifically, in the first year of this study,

* field investigation and interviews were conducted toprovide important descriptive data and documentationof the most effective practices for integratingmicrocomputer instruction into ongoing educationservices for the handicapped. In the second year ofthe study, these "most effective practices" will beimplemented and demonstrated in practicum siteslocated within schools serving the handicapped. Inaddition, a collaborative effort with LIU for thetraining of school personnel at these practicum siteswill take place. Lastly, in the final projectmonths, a large scale collaborative disseminationeffort related to contextual and training requirements necessary for the effective integration ofmicrocomputer instruction into ongoing educationservices for the handicapped will be conducted.

RESULTS The grant is now in Year One. Data is being collected and analyzed. However, we feel that findingsassociated with this grant will concentrate on:

1. The need for leadership and formal structuresthat facilitate efficient planning and decisionmaking at all levels within a district.

III

2. Ways to introduce computers into the schools.

3. Agreement on specific kinds of training to meetthe needs of different types of teachers.

4. The particular skills and competencies thatspecial education professionals ought to have inthe use of computer technology.

5. Strategies for teaching these skills.

FUNDING U. S. Department of Education, Office of SpecialSOURCE Education and Rehabilitative Services/Special Education Programs

RESEARCHER Philip ArcherCoordinator, Office of Institutional

Planning & ResearchNassau BOCESValentines Road & The Plain RoadWestbury, New York 11590516-997-8700 ext. 330

72

ARCHER

STUDY The Impact of Microcomputer Instruction on Handi-capped Children

SUMMARY The purpose of this project was to study: (1) theeffectiveness of microcomputer instruction in promo-ting the learning of handicapped students in a schoolsetting; and (2) the organizational and administra-tive requirements associated with implementing amicrocomputer-based instructional program in a largeschool setting.

The evaluation objectives were to: (1) determine theeffectiveness of different types of microcomputerinstruction on the readiness/achievement of preschooland elementary students according to level of cogni-tive development; (2) determine the effectiveness ofdifferent types of microcomputer instruction on thebehavior of preschool and elementary school studentsaccording to level of cognitive development; (3)determine the effectiveness of different types ofmicrocomputer instruction on the problem-solvingability of preschool students according to level ofcognitive development; (4) determine the effective-ness of different types of microcomputer instructionon the general cognitive ability of preschool stu-dents; and (5) determine the administrative, organi-zational, and logistic requirements of implementingmicrocomputer instruction as associated with studentlearning.

Students in one preschool (ages 3-5), two early ele-mentary schools (5-8 years), and one upper elementaryschool (ages 8-11) received computer instruction forthree 3B-minute periods a week. There were 120 stu-dents in the preschool and 41 in the building'selementary program, 115 and 127 in the two earlyelementary schools, and 196 in the upper elementaryschool. Instruction took place in each school'sLearning Center which housed Apple and Texas Instru-ments microcomputers. Students were randomly assig-ned to three treatment groups according to levels ofcognitive functioning. Treatments consisted of Logo,instructional software (selected on the basis ofstudents Individual Education Plans), and teacher-directed activities (control).

Preschool children were assessed by their teachers onselected subjects of the Bri ance Diagnostic Inven-tory of Erl Develo?ment and t ie Classroom Ben715FInventory, Preschool Form (schaere17-17-EdgFROTIT7Project staff adinialteMindividually the Think It

73

I Through problem solving subtests of Circus andselected subtests of the Kaufman Assessment Batterxfor Children. ElementariMnrits wereTrVen appro-FriarrieVeTs of the Metropolitan Readiness Tests andMetro olitan Achievement Tests in and mathand

iwere raterTrIn-gall egrarers on the Classroom

Behavior Inventory (Schaefer, Edgerton & Aaronson,1978).

RESULTS Although there were significant differences accordingto blocks (levels of cognitive and linguisticfunctioning) there were no statistically significantdifferences according to treatment except for thatbetween the Logo and control groups on the problemsolving test. Logo students scored higher.

It is believed that the necessarily short treatmentperiod (three to four months) contributed in largemeasure to the absence of significant differences.The 1984-85 year treatment period will be at leastseven months so that findings will be more reflectiveof the nature of the treatments rather than of theirintensity.

It was evident that Learning Center staff had hadlittle exposure to Logo and received only rudimentarytraining from project staff. This year Logospecialists offered regularly scheduled training toLearning Center staff in order to insure that theLogo treatment is indeed representative of Logo asconceived by Papert.

The validity of the 1984-85 treatment period assuresus that the findings, whatever they may be, will haveimportant implications for the education of specialeducation students.

Selected students who are representative of thedesignated blocks will be observed regularly by pro-ject staff employing the instrument developed duringthe first project according to Sigel's (1977) Modelof Cognitive Distancing. Content validity of theproject's Microcomputer Observation Instrument wasestablished by relating Sigel's interactions betweenchild/computer and teacher/child/computer. Sigel'soriginal behavioral descriptions, which were derivedfrom parent/child interactions, were videotaped andthen categorized along a continuum ranging from lowto high distance. Inter-rater reliability was alsoachieved by viewing videotaped lessons. Agreement oflevel and category was .80 among four trained raters.In this way the process of learning through computerinstruction can be documented and compared.

7 4

411 FUNDING U. S. Department of Education, Office of SpecialSOURCE Education and Rehabilitative Services/Special Educa

tion Programs

RESEARCHER Philip ArcherCoordinator, Office of Institutional

Planning and ResearchNassau Board of Cooperative Educational ServicesValentines Road and The Plain RoadWestbury, New York 11590516-997-8700 ext. 330

STUDY

SUMMARY

AWAD-EDWARDSENGELHART

Training Methodology and Results Eor a Voice-Controlled Robotic Aid

When we began our research in 1981, there were nohistorical precedents; therefore, no establishedprotocols for teaching individuals to use voice-controlled assistive devices such as a robotic aid.Within a new methodological framework (InteractiveEvaluation) for evaluating prototype assistivedevices, training procedures have been developed,standardized, and researched for a voice-controlledrobotic aid. Training is important to the prototypeevaluation research process because it (a) provides amechanism for systematic collection of user feedback,(b) allows opportunity for depth and breadth of userinput, and (c) is critical to acceptance of anyadvanced assistive device --- particularly computer-ized, interactive devices, such as a robotic aid. Afundamental premise of Interactive Evaluation is thatpotential users are directly involved throughout anassistive device's research, development, evaluation,and diffusion stages.

The Veterans Administration/Stanford University(VA/SU) Robotic Aid Project is the first system toincorporate a human-scale industrial manipulator withsix degrees of freedom (Unimation PUMA-150), a stan-dard microprocessor-based voice recognition unit(Interstate Electronics VOTERM), and synthesizedvoice response (Votrax), and to be driien by high-level software on a Zilog MCZ 1/25 microprocessor,using totally digital, multiprocessor-based algo-rithms.

User training on this system is accomplished in fivesessions which are one to two hours in length.During each session, the user reviews existing exper-tise, acquires new information, practices the lessonsin both structured and unstructured manners, andfinishes the session with a task that helps the userassimilate the entire lesson. While we have trainedusers from age 5 to age 90, our approach is based onandrogogical and pedagogical principals. Key con-cepts include self-directedness, learning readi-ness, immediate applicability, and problem-centeredlearning tasks.

IIIOver 110 users have been trained to use the VA/SURobotic Aid. A 200-page user's manual (containing

73

RESULTS

FUNDINGSOURCE

RESEARCHER

both self-paced and trainer-paced instruction) hasbeen developed and utilized by people with widelyvaried educational backgrounds. These educationallevels range from less than a high school deg'ree toPh.D.s, both technical and non-technical, and M.D.s.Feasibility of training a wide age, educational, andability range has been shown.

Tasks for the robotic aid have been defined andstudied in the following areas:

o personal tasks such as activities of dailyliving (ADL) including cooking, serving, saltshaking;

o recreational tasks such as board games andpainting;

o vocational tasks such as opening file drawers,extracting files, and presenting them;

o therapeutic tasks to improve visual monitoringskills and range of motion therapy.

Proficiency levels for robotic use have been estab-lished based upon these task areas. All userslearned to pick up a cup within one hour of theirfirst introduction to the robotic aid system.Drinking, feeding, serving, cooking, picking up andplacing and fetching tasks for a robotic aid werenamed by 90% of the quadriplegic respondents.

Development of standardized training procedures forinstructing trainers is currently underway. A "peertraining" program is also being developed to matchuser and trainer according to age or personal rehabi-litation experience. Future studies are designed toassess the ability to optimize successful use of therobotic aid by matching a particular training stylewith a particular set of user characteristics.

None Specified

Roger Awad-EdwardsResearch Health ScientistRehabilitation Research & Development CenterPalo Alto VA Medical Center3801 Miranda Avenue, #153Palo Alto, CA 94304415-493-5000

and

K. G. EngeihartResearch Health ScientistRehabilitation Research & Development CenterPalo Alto VA Medical Center3801 Miranda Avenue, #153Palo Alto. CA 94304415-493-5000

BEHRMANNLAHM

STUDY Technology to Assist Young Multiply HandicappedChildren

SUMMARY Two related research projects investigated the use ofcomputer technology to enable young multiply handi-capped children to interact with their environment.

The first study addressed the question of the minimaldevelopmental levels (e.g. cognition, language) re-quired for a child to successfully activate a switchcontrolled computer. The methodology included compu-ter generated voice prompts to activate a switch.Response data were collected directly by the compu-ter. Developmental ages, as recorded on studentEarly Learning Accomplishment Profiles, were col-lected. A multiple linear regression was used toanalyze the relationship of developmental ages toachieving pre-determined mastery criterion forresponse time.

The second study evaluated the role and behaviors ofthe teacher, computer prompts, and the child, and thesequence of behaviors required to teach the skill ofactivating the computer with a switch. The behaviorof all three were coded based on videotapes of teach-ing sessions. A lags sequential analysis was per-formed on the data.

RESULTS In the first study cognitive age was found to bestpredict mastery, with a developmental level of 11.4months the minimal age required to master the neces-sary skill. This study shows that very youngchildren can be taught to interact with a computersystem. It shows the potential for using computertechnology for early intervention in the area ofenvironmental control for young multiply handicappedchildren.

In the second study, the findings indicated that thechild achieved the skill without interaction with theteacher. Despite the achievement of the target skillby the child, teacher intervention continued. Thissecond study raises numerous questions about the roleof the teacher in that teaching process. Additional-ly, it illustrates that more thought must be given tothe presentation sequence of the learning task toachieve efficient learning on the computer.

411FUNDING None SpecifiedSOURCE

0 RESEARCHER Michael M. BehrmannAssistant ProfessorDepartment of EducationGeorge Mason University4400 University DriveFairfax, Virginia 22030703-323-2541SpecialNet: Behrmann or GeorgeMason.Univ

and

Liz LahmThe COSMOS Corporation1735 Eye Street, N.W.Suite 613Washington, DC 20006202-728-3939

80

e BENNETT

STUDY AERA Editor-at-Large Program Liaison for Microcom-puter Research in Special Education

SUMMARY The AERA Editor-at-Large Program works to identifyand publicize quality research in a variety oftopical areas related to education. As the ProgramLiaison or "Acquisition Editor," for Research onMicrocomputers and Special Education, I welcomehearing from researchers about their work in thisarea.

RESEARCBF? Randy BennettDirectorTechnology LabEducational Testing ServicePrinceton, New Jersey 08541609-734-1101

e BRINKERDEMI

STUDY Handicapped Infants Exploration of the World:Building Persistence at an Early Age

SUMMARY Contingency intervention is an approach to teachinghandicapped infants that their behavior controlsinteresting environmental events, thereby buildingthe motivation for exploring that environment. Themicrocomputer is used to: (1) control contingenciesbetween behavior and consequences; (2) collect per-formance data to determine if infants are learning orhave already mastered to relationship between theirbehavior and environmental events; and (3) providesummaries of progress so that teachers can decide tochange the complexity of the environment to be ex-plored. From a variety of theoretical perspectives,the process of learning to control simple consequen-ces provides the foundation for human adaptation.

A review of research results demonstrates: (1)different patterns of learning by handicapped infantswhich characterize different styles of exploration;(2) significant development of moderately handicappedinfants receiving contingency intervention; and (3)the feasibility of using contingency intervention asa process assessment tool for, severely/profoundlyretarded preschoolers for whom standardized measuresdid not provide meaningful information.

Currently, a small business, Contingency Software,Inc., has been funded by the National Institute ofChild Health and Human Development to develop contin-gency intervention software for use by teachers inearly intervention projects. Site selection criteriaand project design for national field testing will bediscussed.

RESULTS The contingency intervention research demonstratesthat, given the appropriate environment, handicappedinfants as young as 3 months of age can discover thatthey control events. Moreover, once they have dis-covered a simple control rule they will not persistindefinitely emitting behavior and prcAucing con-sequences. This breakdown of behavioral control canbe explained in terms of cognitive models of habitua-tion. Behavioral control and renewed interest on thepart of the handicapped infant are re-establishedwhen a new contingency problem is presented. Thuseven for handicapped infants Papousek's observationsseems to apply: The primary motivation is to explore

82

FUNDINGSOURCE

RESEARCHER

new problems and solve new contingencies. The impli-cations of these findings could fundamentally alterour notion of behavioral control and its older sib-ling compliance. The absence of behavioral controlmay reflect the teacher's failure to detect the in-fants solution of one problem and to provide a newmore challenging problem.

National Institute of Child Health and Human Develop-ment

Richard P. BrinkerDirector of Early Intervention1640 West Roosevelt RoadChicago. Illinois 60608312-996-1567

and

Richard DeniRider College2083 Lawrenceville RoadTrenton, New Jersey 08648609-896-5096

83

I

STUDY

CARNINEWOODWARD

Computer Simulation Instruction on Health Facts andProblem*Solving Strategies for Learning Disabled HighSchool Students

SUMMARY The purpose of the study was to investigate theeffectiveness of a health computer simulation inteaching high school mildly handicapped studentsbasic health concepts and strategies. Thirty stu.tdents from a Eugene high school were randomly assign*ed to one of two groups: (1) conventional healthinstruction and (2) a combination of conventionalhealth instruction and the simulation. Each day, allstudents received the conventional instruction as onegroup for 20 minutes. For the remainder of thelesson, the conventional instruction group worked ata variety of extension and review activities. Eachstudent in the simulation group worked at a microcom*puter with Health Ways.

After 12 days of instruction, each student was giventwo posttests and an attitude survey. The HealthWays Nutrition and Disease Test was a 30.item testdesigned to measure retention of basic health factsand concepts from the written curriculum. The HealthWays Diagnosis Test measured the student's ability toprioritize changes in a person's health habits basedon diet, heredity, current disease, and other essen<tial health related information. The Nutrition andDisease Test was given again two weeks later as ameasure of retention of the fact and concept information.

Both posttests were also given to a random selectionof high school students from health classes. Theirscores were compared with those of the two groupsthat participated in the study. Analyses of variance(ANOVA) and t*tests were performed on the data.

A properly developed computer simulation can producemore sophisticated problem*solving behaviors (relatedto analyzing lifestyles) in learning disabled stu*dents than is found in non*learning disabled studentswho are enrolled in regular health classes. Theselearning disabled students are now tutoring regulareducation students on health promotion, using thesimulation. Wedding technology and analytic instruc*tional design can produce substantial learning ofcomplex skills in mildly handicapped students.

S

S

RESULTS The major purpose of this study was to determinewhether or not a computer simulation used in combins.ttion with a direct instruction approach to a healthcurriculum (simulation group) is more effective thanthe direct instruction method alone (conventionalgroup). In the main analysis, these two methods arecompared on two measures. A supplemental analysisfurther compares the performance of these two groupswith students randomly selected from regular highschool health classes on the same two measures.

Three separate 2 x 2 analyses of variance (ANOVAs)were performed on scores from the different sectionsof the Health Ways Nutrition and Disease Test.Analyses included: (1) the section of the test notreinforced by the Health Ways simulation; (2) thesection of the test TeTiTro-rFe-aby the simulation; and(3) the total test. The analyses of mean scores ofthe two groups on the Nutrition and Disease Testgiven one day following and two weeks after instruc.*tion reveal a significant main effect for theinstruction method on the total test score and onthose items in the test that were reinforced by theHealth Ways simulation (a < .03 and E < .01), respec.*trieTy-). The section where items were not reinforcedwas n ar significance (< .06). There was no signifi.ecant interaction between instructional method andtime of testing.

Five t.*tests were performed on the posttest scores ofthe Health Ways Diagnosis Test. Scores for the con.*ventional and simulation groups were compared in thefollowing areas: (1) prioritizing alone; (2) stressmanagement; (3) identifying health problems andmaking correlated changes; (4) total test score with.*out stress as a factor; and (5) total test score.The findings reveal significant differences (< .001)in all areas.

In supplemental analysis, a one-way analysis ofvariance (ANOVA) was used to compare the test perfor<mance of the conventional and simulation groups withstudents from regular health classes who did notparticipate in the study. Again, scores from eachsection of the Health Ways Nutrition and Disease Testand the Health Ways Diagnosis Test were analyzed. Onthe Nutrition and Disease Test, only the sectionwhere items were reinforced by the Health Ways simu.elation showed a significant difference (a < .01).Total test scores were near the significance level oa< .07). All sections of the Diagnosis Test showedsignificant differences (a < .001).

85

IIIFUNDINGSOURCE

U. S. Department of Education, Office of SpecialEducation and Rehabilitative Services/Special Educa<tion Programs.

RESEARCHER Douglas CarnineAssociate ProfessorDivision of Teacher EducationCollege of EducationUniversity of OregonEugene, Oregon 97403503.e485.e1163

and

John WoodwardResearch AssociateDivision of Teacher EducationCollege of EducationUniversity of OregonEugene, Oregon 97403503485.e1163

s

0

CARTWRIGHT

STUDY Computer Assisted Instruction in Teacher Education:A Full Length Course

SUMMARY In addition to its main campus, The PennsylvaniaState University maintains 17 other campuses wherefreshmen and sophomores complete the first 2 years ofundergraduate work. Some majors, though, do not havefaculty at all campuses and students must wait untiltheir junior year to take their first courses intheir majors. Although there are not educationalfaculty at some campuses, there is a great need tobegin instruction in the field of teacher educationbefore the third year of a 4-year program.

In response to that problem, the University has deve-loped, evaluated, and implemented a series of modulesand an entire 3-credit teacher education course whichis offered completely by microcomputer. In addition,an innovative microcomputer-based procedure was deve-loped to evaluate the extent to which students andinservice teachers have achieved certain. competen-cies.

EDUCATING SPECIAL LEARNERS, a full length 3-semesterhour course, is given by microcomputer as a "standalone" course, All instruction is given by the com-puter through a series of computer modules. Themodules are easily transportable to other institu-tions and use the following hardware: Apple IIseries (II+, IIc, Ile) and the IBM PC series (PC,PCXT, PCjr). For the past 2 1/2 years the course hasbeen used successfully on 13 Penn State campuses.This computer-assisted instruction course can also betaken through Penn State's continuing education andcorrespondence programs.

A series of studies, conducted primarily by ProfessorPatrick Schloss, have focused on the following as-pects of computer-assisted instruction: location ofquestions and highlights as a variable in computer-assisted instruction; efficacy of various ratios ofquestions and highlights to text; higher cognitiveand factual questions; placement of questions andhighlights as a variable in influencing theeffectiveness of CAI; and focus of control of compu-ter-assisted instruction. These studies have beenconducted using the EDUCATING SPECIAL LEARNERSmodules.

RESULTS Evaluation of the course was based upon three

87

FUNDINGSOURCE

RESEARCHER

criteria: cost efficiency, academic achievement, andstudent acceptance. These three criteria have beenmet and exceeded. Evaluations were carried outinternally at Penn State and by other institutions ofhigher education implementing the course. Currently,the course is in use in approximately four dozeninstitutions in the United States and Canada.

None Specified

G. Phillip CartwrightHeadDivision of Special Education

and Communication DisordersThe Pennsylvania State UniversityUniversity Park, Pennsylvania 16802814-865-6072

CAVALIERMINED

STUDY A Bladder Sensor for Persons with Urinary 'neon.*tinence

SUMMARY The goal of this 3-eyear project is the developmentand field testing of a portable ultrasonic bladdersensor to aid in the toilet training of incontinentpersons and to provide increased independence forpersons who have permanently lost the ability tocontrol their bladders. The aid will continuouslymonitor the volume of urine in a person's bladder andthen provide a subtle auditory, visual, or tactilesignal when a threshold volume is reached.

The project is a collaborative effort of Associationfor Retarded Citizens of the US (ARC/US), the NASAUltrasonics Team at the Langley Research Center, theNASA Biomedical Applications Team at the ResearchTriangle Institute, and the Medical College ofVirginia (MCV). The project is divided into fourmajor phases. During the first phase, ARC/US consoleidated all available research information on urinaryincontinence in the various handicapped populationsinto a computerized database which will be periodictally updated throughout the project. In addition,this phase involved cooperations with MCV to deter.*mine the anatomical/physiological parameters forproper positioning of the bladder sensor. The secondphase will involve the development and in-ehouse test.*ing of first a rack.emounted prototype and then aportable version.

During the third phase, the portable aid developed inphase two will be systematically evaluated in clientservice settings. This behavioral research will takeplace in the Dallas Independent School District andin programs of the ARC-ePeninsula in Virginia togather information: (1) on the durability, reliabil.eity and flexibility 'f the prototype; (2) on theutility of the sensor in providing increased indepenedence in toileting; (3) for guidelines on the approepriate use of the sensor; (4) for credibility of thesensor's effectiveness and acceptance by researchersand practitioners; and (5) for the base for eventualmarketing. The final phase will involve the prepara.*tion, dissemination, and utilization of the projectinformation and reports.

The anticipated implications of this research are:

FUNDINGSOURCE

RESEARCHER

S

1. More efficient and effective toilet trainingprocedures for populations currently unable tobe toilet trained using traditional techniques.

2. Greater independence and improved self-esteemfor incontinent persons with a variety of handi-capping conditions.

U.S. Department of Education, National Institute forHandicapped Research

Al Cavalier.Director, Bioengineering ProgramAssociation for Retarded Citizens of the US2501 Avenue JArlington, Texas 76006817-640-0204

and

Beth MineoProject DirectorAssociation for Retarded Citizens of the US2501 Avenue JArlington, Texas 76006817-640-0204

CAVALIERMINED

STUDY Technology to Enhance Special Education: Remediationof Problems in Logical Thinking and Memory

SUMMARY Memory problems are believed to exist in the largemajority of persons with mental retardation andlearning disabilities. If deficiencies remain uncor.*rested they compound higher level areas of function.*ing and frustrate remediation attempts.

The instructional systems under development have beendesigned for use with two popular personal computers.The systems combine techniques for teaching efficientcognitive strategies with the unique characteristicsof computers in logical analysis, memory, and motiva.*tion. Through an interactive format, the systemfirst assesses the degree and type of memory processdeficiencies and then provides individualizedremedial instruction appropriate to the assessmentresults. The system also maintains a record of eachstudent's performance for review by parents andteachers.

Field testing of the instructional system will beconducted in an attempt to answer five research ques.*tions:

1. Will the computerized version of the assessment/instruction procedure yield data similar tothose derived from previous laboratory andclassroom research;

2. Does the assessment identify memory problems;

3. What is the nature of memory problems evidencedby individuals with mental retardation orlearning disabilities, and are there any differ.*ences between the two groups;

4. Do the instructional techniques employed assistin remediating the rehearsal deficiencies iden<tified in the assessment; and

5. Do students generalize the use of strategiestrained directly to instances for which theyreceived no training?

Approximately 60 subjects matched for chronologicalage will serve as subjects. Twenty subjects withmental retardation, 20 with learning disabilities,

and 20 nonhandicapped students will participate.Half of the subjects in each group will serve ascontrols while half will receive intervention withthe computer-based instructional package.

RESULTS The results of this research should lay a foundationfor future work concerned with computer-assistedremediation of memory deficiencies. It will allowconclusions to be drawn regarding the appropriatenessand effectiveness of computer-based interventions ofthis type.

FUNDING U. S. Department of Education, Office of Special Edu-SOURCE cation and Rehabilitative Services/Special Education

Programs

RESEARCHER Al CavalierDirector, Bioengineering ProgramAssociation for Retarded Citizens of the US2501 Avenue JArlington, Texas 76006817-640-0204

and

Beth MineoProject DirectorAssociation for Retarded Citizens of the US2501 Avenue JArlington, Texas 76006817-640-0204

STUDY

SUMMARY

FUNDINGSOURCE

RESEARCHER

CLYMER

Instructional Development of Communication TrainingMaterials for the Deaf

The use of educational technology is an integral partof the instructional development (ID) activities atthe National Technical Institute for the Deaf (NTID).My work focuses on several areas:

1. Microcomputers as a Productivity Tool to ImproveID.

In what ways can the new technology improve thesystematic development of special education curriculum? I am interested in applications of microcomputers by expert instructional developers and subjectmatter specialists to improve the efficiency andeffectiveness of the instructional development process. The use of various productivity software isbeing investigated to improve the analysis, design,and evaluation stages of ID.

2. Authoring Languages.

Can authoring languages really facilitate thedevelopment of effective CAI? I have been investigating different types of authoring "tools" and haveattempted to determine which are the most appropriatefor use by classroom teachers at NTID.

3. Interactive Video.

Interactive video is an ideal instructional formatfor providing hearing impaired students with receptive communication training. NTID has used the DAVIDsystem with success for speechreading and sign language instruction. Additional work is being completed on different subject areas as well as development of interactive videodisc applications. Programdevelopment using authoring languages is also beinginvestigated.

None Specified

E. William ClymerInstructional DeveloperCommunication Research DepartmentNational Technical Institute for the Deaf/RITPost Office Box 9887Rochester, New York 14632-0887716-475-6894

I

I

EGOLFBRUNO

STUDY Use and Evaluation of Technology to Promote Communi<cation in Nonspeakers

SUMMARY The purposes of this project are to: (1) evaluate theinteraction patterns between nonspeakers, using elec.etronic communication aids, and their communicationpartners; (2) determine the interactive strategiesthat can enable nonspeakers to become more effectivecommunicators; (3) teach nonspeakers these effectivestrategies; (4) conduct inservice training sessionsteaching interactive skills to nonspeakers; (5) deve<lop a graduate curriculum in the nonspeaking or aug*mentative areas; and (6) disseminate the results of*he research to professionals, parents, and nonespeakers.

The implications of this research deal with therealization of educational, vocational, and socialenhancements by nonspeakers using electronic communi*cation aids. The research is designed to facilitatethe realization by teaching nonspeakers effectivecommunication strategies.

FUNDING U. S. Department of Education, Office of SpecialSOURCE Education and Rehabilitative Services/Special Educa,*

tion Programs.

RESEARCHER Donald B. EgolfAssociate ProfessorDepartment of Communication1117 CLUniversity of PittsburghPittsburgh, Pennsylvania 15260412-*624*6763

and

Colleen A. Haney BrunoSpeech Pathologist/Research CoordinatorDepartment of Communication1104 CLUniversity of PittsburghPittsburgh, Pennsylvania 15260412.e624.e6763

FITZGERALD

STUDY Classroom Behavior Record: An Interactive VideodiscTraining Program in Classroom Observation Skills

SUMMARY The Classroom Behavior Record (CBR), a classroomobservation code, was developed in 1977 as a means togather data of overt behaviors displayed in schoolsettings by children referred to a psychiatricclinic. CBR is a 32 variable code used in 6secondtimed intervals to record data on a target child ofinterest and his/her peers within the same situation.Two versions of this system now operate in Iowa: (1)data collected in a Datamyte or TRS-100 for processing via the University's PRIME computer system orvia the microcomputer program, and (2) data collectedon computer scoring scan sheets and processed via theExam Service. Classroom behavior data has been collected on children with behavior and learning disorders over the past 8 years and used in researchstudies as well as clinical treatment evaluation.

We have now developed an interactive computervideodisc training system to provide complete training onthe CBR and observational procedures. The trainingside of the videodisc contains motion samples ofclassroom behavior, narrative descriptions, stillframe examples and nonexamples of each behavior,glossary information, and practice sessions ultilizing the Power Pad with a coding score sheet. Thepractice side of the videodisc provides extendedmotion samples of classroom behavior which can beused for practice and, we hope, reliability testing.

When the effectiveness of the Interactive videodiscprogram is established, we expect to have a readilyavailable, efficient procedure for training classroomobservers. Presently it takes 40 to 60 hours oftrainer time to bring an observer to the desiredlevel of proriciency and reliability. We expect thevideodisc program will reduce the time the traineespends becoming an observer, and will nearly eliminate the need for an active trainer. The programwill more adequately meet individual needs; there areoften wide variations in tt.e time it takes individuals to learn the code, in their abilities to codeefficiently, and in their motivational commitment tolearn. We further expect that the videodisc willhelp standardize the training and reduce observerdrift over time. Establishing the effectiveness ofthis videodisc program as a "stand alone" trainingpackage will greatly increase its portability to

41/school systems, teacher training institutions, andresearchers throughout the nation.

FUNDINGSOURCE

RESEARCHER

Field test data will be compiled during the 1985-86year on the relative efficiency and effectiveness ofthe videodisc program for teaching observation skillsto an acceptable level of reliability. Training inobservation skills will be incorporated into thegraduate course of instruction in special educationand school psychology and placed in field test districts.

State of Iowa Part B, E.H.A. Discretionary Funds;University of Iowa Video Production Grant

Gail FitzgeraldDirector of Educational ServicesChild Psychiatry Service and LecturerUniversity of IowaCollege of Education500 Newton RoadIowa City, Iowa 52242319-353-3528

FITZGERALD

STUDY Computer Applications for Students with Behavior andLearning Problems

SUMMARY In a 2year research and dissemination project,activities were undertaken to investigate specialmicrocomputer applications in the curriculum forstudents with behavior and emotional disorders, andto provide related skill training to teachers.Research activities have focused on four areas.

1. When given a choice of classroom rewards, howpopular is computer free time in comparison toother typical classroom rewards?

On a forced choice reinforcement questionnaire, andby actual choice, students with severe behavior andemotional disorders were found to select computeractivities significantly more often than comparablefree time or edible reinforcers. A followup studywas carried out to determine the relative strength ofcomputer activity contingencies on classroom behaviorand work productivity when contrasted to a no rewardcondition and a free time/sticker choice condition.

2. How effective is computerassisted instructionwith students with severe attentional difficulties?

This study was designed to contrast computerassistedinstruction with traditional, drillandpracticeinstruction among a sample of children experiencingattentional difficulties. Each student was given anindividualized spelling test to form equivalent worklists for weekly spelling study. Half of the wordswere studied via a CAI program and half in a traditional structured paperpencil drill format.

3. How effective are computerassisted activitiescompared to traditional activities for teachingstudents problemsolving and impulse controlskills?

A curriculum unit was developed to provide trainingon problem solving skills using self instructionalphrases, opportunities to practice with "safe"activities, and applications of the skills to socialsituations. Adolescent behaviorally disordered students in public schools participated in the study.While the teaching and application activities wereconsistent for both study groups, half the students

practiced with computer activities (CAI) using simulation software and half with comparable noncomputeractivities (NC). Comparisons were made of theirperformances on the MeansEnds Test, Kendall SelfControl Rating Scale, and the Porteus Mazes.

4. How do student cooperative work skills on computerbased projects compare to such skills onsimilar, noncomputerbased activities?

Students in a psychiatric hospital setting worked inpairs on cooperative learning activities. A comparison was made of their work skills and interactionalbehavior while involved with computerbased and noncomputerbased projects. Activities were matched asclosely as possible and included cooperative solvingof mazes, writing and illustrating stories, creatingsound tracks for plays, and generating graphicdesigns. Analogue rating scales were completed byeach teacher during the activity periods.

RESULTS 1. Students were found to be ontask more of theirwork time (88% compared to 55% and 63%) and to produce more work (130 correct answers compared to 95and 77) under the computer reinforcement conditionwhen contrasted to a no reward condition and a freetime/stickers condition. These studies suggest thatthe contingent use of computer free time may be acondition and may be a popular and powerful reinforcer for studencs with behavior and emotional difficulties.

2. Findings over a 5week period indicated that thetwo practice methods were equally and significantlyeffective in increasing the number of words learnedover a no practice condition. This study suggeststhat children noted for having problems sustainingattention are not overly distracted by computer gadgetry and software features and can learn through CAIat a level comparable to more traditional methods.By having children spend some time on tne computer,teachers will increase available time for individualized instruction.

3. Mixed results were found. The CAI group showedsignificant improvement on the Porteus Mazes andSelfControl Rating Scale, and the NC group demonstrated greater alternative thinking skill abilitieson the MeansEnds Test. This study demonstrates thatselfregulatory behavior can be enhanced by directinstruction and that computer activities can effectively be incorporated into such instruction.

FUNDINGSOURCE

RESEARCHER

4. The pilot study showed an increase in taskinvolvement behaviors when working on computerbasedactivities. Mixed results were found on other behaviors being rated. A replication of this study in asevere BD public school classroom is currently inprogress, with systematic observation data beinggathered weekly by trained observers. These studieswill evaluate the potential of using CAI activitiesto encourage cooperative behavior skills necessaryfor appropriate social functioning.

State of Iowa Part B, E.H.A., Discretionary Funds

Gail FitzgeraldDirector of Educational ServicesChild Psychiatry Serviceand Lecturer

College of EducationUniversity of Iowa500 Newton RoadIowa City, Iowa 52242319-353-3528

eSTUDY

SUMMARY

FUNDINGSOURCE

RESEARCHER

FRIEDMAN

Eye Control: Using Vision as an Effector as well asa Sensory System

Although vision is one of our primary sensory sys-tems, our eyes have always been used to communicateas well. Where we look during a conversation is oneof the cues for nonverbal coordination of turn-takingduring conversations. Similarly, others can "read"something about our emotional and/or attentionalstate by observing the size of the pupils of oureyes. Modern technology allows us to control compu-ters and, through computers, to control robotic andenvironmental control systems with our eyes.

In special education, eye control of computer commun-ication systems is a method of choice for manyseverely physically handicapped children. It allowsthem to communicate at relatively high speed forextended time periods without excessive fatigue. Onecommercial "spin-off" of eye movement research is theEyeTyper, a noncontact, eye-gaze-controlled keyboard.Some current research is directed towards using theeyes for "mouse" or "joystick" control of computersoftware. Other research is aimed at direct eye gazecontrol of target selection for robotic systems. Thelatter research promises to enhance manipulativecontrol of the environment for the severely phys-ically disabled. The use of eye gaze control devicesmay lead to subtle, unconscious changes in eye move-ment patterns.

In special education today, it is the communicationneeds of physically handicapped children that arebeing addressed by eye movement analyzing systems.In the future, the communication needs of severelyemotionally disturbed children may be aided as well.One can consider modifying toys and appliances (tele-vision sets, doors, etc) to respond to being lookedat as a first step toward interactive communicationby those who are unresponsive to normal communicationchannels.

Personal resources and venture capital.

Mark B. FriedmanResearch EngineerRobotics InstituteCarnegie-Mellon UniversityPittsburgh, PA 15213412-578-7692

STUDY

GALL

The "Prairie Post": A Microcomputer-Based ElectronicBulletin Board Serving Rural Special Needs Popula-tions

SUMMARY The objective of this field-based research projectwas to provide an opportunity to examine the efficacyof a cost-effective microcomputer-based electronicbulletin board system in helping to overcome thespecial communication difficulties encountered byeducators serving special needs populations in ruralareas. The venue for the project became theeducational jurisdictions presiding in the ruralregion of Southwestern Alberta (Canada) bounded inthe south by the American border and the west by theRocky Mountains, and covering an area of approxi-mately 18,000 square miles. The academic base forthe project was the University of Lethbridge, and themandate for the project was obtained from theDepartment of Advanced Education (Government ofAlberta). It was identified as the Rural SpecialEduce-kon Outreach Project (RSEOP).

The Prairie Post is a computer-based informationsystem -- commonly called a "bulletin board." Infor-mation is shared through the use of computers. Al-though it is not a new concept to the zealots incomputer circles, it is seen as an innovation to manyeducators.

Conceived in the fall of 1983, the Prairie Post waslaunched in the summer of 1984 as a pilot of itsparent -- Rural Special Education Outreach Project.The intent of the Prairie Post was to place teachersand administrators in rural areas in contact with thelatest developments in special education. In addi-tion, this relatively new medium would also allow thesame educators to share information amongst them-selves through the convenience of electronic net-working.

Information regarding specific components of thePrairie Post, and appropriate data regarding modifiedinteraction patterns resulting from the project, canbe obtained from the author. The major componentsinclude these primary objectives:

1. To develop a regional physical network of hard-ware units, which would become the RSEOP BulletinBoard, and to establish a communications networkamong among them.

2. To share these units in a pilot phase of the

project between selected representatives of theteaching and administrative members of the schoolsystems within the rural region of SouthwesternAlberta, and to eventually extend these to parentsand special needs children.

3. To "broker" information services available oninformation utilities such as SpecialNet and BRS-After Dark on behalf of the RSEOP pilot projectparticipants, and to establish appropriate subscrip-tion contracts with these utilities.

4. To encourage the development of a variety oftelecommunications applications through the RSEOPnetwork, including electronic mail between partici-pants and, externally, to the information utilities,consultation services among participants, universitypersonnel, and the information utilities; and elec-tronic conferencing between system participants onmatters of common interest.

RESULTS The following schematic of on-line user accessprovides one sample of the callers access' frequency.

3/25/85 BULLETIN BOARD USAGE

160

140

120

100

CALLS 80

60 Each represents 1caller

40 Average connect time =9 minutes

20

0

HOURS

s RESEARCHER Robert S. GallEducational PsychologistProfessor of Education andAssociate, National Institute onMental Retardation

The University of Lethbridge4401 University DriveLethbridge, AlbertaCanada T1K 3M4403-329-2461SpecialNet: LETHBRIDGE

1.03

HAGEN

STUDY Identifying and Using Authoring Systems for SpecialNeeds Children

SUMMARY This study was designed to identify and use two "off-the-shelf" authoring systems useable with specialneeds children for language development; one in eachof two classifications: simple-to-use mini authoringand more complex major authoring systems.

Phase One: Identify criteria for selection:

Considerations for Mini Authoring:

1. Useable with single disk drive Apple II Systems.

2. Cost

3. User Control

4. Large Type

5. Fixed Format/Menu Driven for ease of use byteacher, for simple Drill and Practice Materials.

Considerations for Major Authoring System:

1. Useable with single disk drive Apple II Systems.

2. Cost

3. User Control (no timing loops)

4. Large Type

5. Graphics Ability

6. Color

7. Speech Output

8. From full keyboard to single switch access

Phase Two: To implement use with two hearingimpaired and six learning disabled students.

The methodology for choice was to examine commer-cially available authoring systems existing at thetime (1981-1982) and to use those that came closestto the criteria standards.

0 The methodology for implementation was an independentstudy time extracted from regular study periods toinclude a minimum of 20 minutes per day. Measures ofeffectiveness included grade improvement, attitudeand motivation of students as well as the scoring ofactual courseware lessons.

Two authoring systems were identified: MICROQUEST asthe mini authoring system meeting all of the criteriafor selection; and E-Z PILOT as the major authoringsystem meeting (in its original form) 6 of the 8selection criteria.

RESULTS Implementation of these authoring systems wasextremely successful with the two deaf students. Ameasurable improvement in vocabulary usage existedfor all eight of the students. Improved vocabularyunderstanding with usage both written and oral wasdocumented through grade improvement and actualcourseware scoring. Motivational impact showed moreaggressive participation in regular classroom activi-ties as well as improved independent study habitsusing the vocabulary support software preparedthrough the authoring systems.

The use of authoring systems, such as the two used inthis project, demonstrated the effectiveness of inde-pendent visual drill and practice for the two deafstudents. The successful but less dramatic resultsin direct vocabulary improvement for the LD studentscreated the desire to add speech to the softwarepresented to the learning disabled students. Thisdesire led to contact with the author of E-Z PILOTand the eventual inclusion of speech output throughthe ECHO II speech synthesizer. It also led to theuse of the Hartley Cassette Control device and theability to add tape recorded human speech to thecourseware.

Based on the effectiveness of this project and otherslike it, the desire to make curriculum specificcourseware created on E-Z PILOT available to otherpopulations, such as severely physically disabled andthe mentally disabled, arose. This led to a specialversion of E-Z PILOT that allows single switch entrythrough the game control port, either through touchsensitive alternative keyboards or the addition ofthe Switch Interface for common mini-plug switches.

The overall implication was the creation of a multi-sensory authoring system using low cost commerciallyavailable products: E-Z PILOT, ECHO II, Chalkboard,Hartley Cassette Control Device, Switch Interface,

105

FUNDINGSOURCE

RESEARCHER

etc., all with the intention of making microcomputer-based learning materials accessible to all disabil-ity groups.

Private

Dolores and Bud HagenPublisher, CLOSING THE GAPPost Office Box 68Henderson, MN 56044612-248-3294SpecialNet: CTG

106

STUDY

HAMLETT

An Investigation of the Feasibility of a Microcom-puter System for Developing Scanning and SelectionSkills in Severely Handicapped Students

SUMMARY Electronic scanning communicators may be the mostefficient alternative communiation system for indi-viduals with severe physical limitations. Scanningcommunication aids reduce the number and precision ofmotor movements required for a child to initiate com-munication. A primary disadvantage, however, is thatthe scanning process requires more sophisticatedskills than direct selection. The user must be ableto visualize his/her goal, to anticipate the direc-tion of the cursor, and to anticipate the arrival ofthe cursor to the goal. Scanning may also negativelyaffect motivation because of the slowness of thetechnique. Some physically impaired children, parti-cularly those who also have significant cognitivedeficits, may require extensive training. Verylittle research has been reported which addresses theissue of effective training strategies for teachingchildren with severe handicaps to use electronicscanning devices.

The current research addresses the effectiveness of amicrocomputer video game designed to train nonvocalseverely physically handicapped students to make scanand selection responses similar to those needed foroperating linear scanning augmentative communicationaids. The video scanning and selection game systema-tically shapes scanning on more complex screens byslowly increasing the number of boxes presentedsimultaneously on the screen.

The primary study evaluates the effectiveness of thevideo game in increasing the number of correct scanand selection responses for each subject using a

single subject multiple problem across subjectdesign. Periodic probes are used to examine thedegree to which scanning and selection responsestransfer to a standard three symbol light scanningcommunication aid. A related study, using a Latinsquares design, examines the use of three differentcursor types by young (3-year-old) nonhandicappedstudents.

RESULTS The most significant implication of this research is

for communication training for severely handicappedchildren with significant cognitive deficits. Manyof these children who learn to operate a rotary

FUNDINGSOURCE

RESEARCHER

scanning device have difficulty making a transitionto more versatile and powerful linear scanning communication aides. One possible reason for this difficulty is that a rotary scanner provides additionalcues to the child regarding the direction and speedof the cursor. Using a microcomputer for trainingwould allow the use of intermediate training stepswhich retain the additional cueing of the rotaryscanner while using the format of linear scanning.In addition, the flexibility of the microcomputerallows easy development of game formats to increasestudent motivation and interest during training.

Spencer Foundation

Carol HamlettResearch AssociateDepartment of Special EducationBox 45 Peabody CollegeVanderbilt UniversityNashville, Tennessee 37203615-322-8070

108

HASSELBRING

STUDY A Chronometric Analysis of the Effects of ComputerBased Drill and Practice in Addition and Subtraction

SUMMARY Chronometric analysis has been used to verify a "MinModel" for solving single digit addition problems.In this model, a child adds two numbers, m and n, bysetting a counter to the maximum value for the twoaddends (m,n), and then incrementing it the samenumber of times as the minimum value of the twoaddends (m,n). Using chronometric analysis, Groenand Parkman (1972) showed that most children have notmemorized their basic fact tables and do not usereproductive memory to recall these facts, butrather, use reconstructive processes to generateanswers using the Min Model.

The purpose of our current research is to determineif computerbased drill and practice can be used toenhance a child's memorization of the basic additionand subtraction facts, thus reducing the child'sdependence on the more primitive reconstructive orcounting process.

In the current set of studies, we provided mildlyhandicapped students with daily computerbased drilland practice in the form of arcade games over a 20 to30 day period.

RESULTS The results of these studies indicate that (1) almostall students increase their rate of correct responding as a result of the drill and practice; however,(2) few students move from the use of reconstructiveprocesses to reproductive processes.

The conclusion drawn from these studies is that computerbased drill and practice in the form of arcadegames does not assist a student in moving fromreconstructive to reproductive processes. However,if a student is using reproductive processes, thenthe use of drill and practice is effective for increasing the student's rate of correct responding.

The findings indicate that computerbased drill andpractice has a positive effect on increasing the rateof reconstructive recall of basic math facts but doesnot, in itself, cause students to use reproductiverecall of facts. Thus, drill and practice isineffective as a tool for memorizing basic additionand subtraction facts and should not be used for thatpurpose.

109

FUNDINGSOURCE

RESEARCHER

Computer technology has allowed us, for the firsttime, to engage in micro level learning research in aclassroom setting with large numbers of students. Weare able to analyze student learning processesthrough automatic data collection that heretoforecould only be done in laboratory settings. Themicrocomputer offers the field of special educationone of the most powerful research tools everdeveloped.

None Specified

Ted HasselbringAssociate ProfessorBox 328Peabody College of Vanderbilt UniversityNashville, Tennessee 37203615-322-.8186SpecialNet: PCVUSPECIALED.

HAYNES

STUDY Effects of Computer.eAssisted Instruction on DisabledReaders' Metacognition and Learning of New Words

SUMMARY Background. Research indicates that one of thecommon features distinguishing successful from lesssuccessful learners is their level of metacognitiveactivity, including the ability to monitor and regu.elate their own processes for learning. The researchliterature on computers in instruction tends to focuson immediate learning. More research is needed onthe effects of computeneassisted instruction on thelearning process itself. This study examined theeffects of CAI on the metacognition of disabledreaders by examining their ability to predict theirachievement level on similar vocabularpedevelopmentactivities delivered with computers and with a tradi.etional practice format.

Procedures. A computer program was developed toteach the spelling, definition, and correct sentenceusage of a list of words. A series of paper.eand.*pencil activities was developed to parallel the corn.*puter program as closely as possible. Four observerswere trained to collect data for the group receivingthe traditional practice method on the number oftrials students performed, the types of trials per.*formed (spelling, definition, or sentence usage), andthe amount of time spent practicing the words. Thecomputer program automatically collected similar datafor that group of subjects.

Subjects for this study were 32 disabled readersenrolled in a summer reading clinic. Subjects ineach group practiced five words using their assignedexperimental condition for up to 15 minutes for fourconsecutive days. Following each practice periodthey were asked to predict how many words they wouldget right on a test. An oral test was then adminis.etered to each student. On the fourth and final dayof practice, subjects were asked to predict how manyof the words they would remember in a week. Thefollowing week the same oral test was administered toall subjects.

RESULTS T.ttests and correlations were used to measure differ.*ences between the computer and traditional groups,and to measure the relationships between certainvariables. Table 1 summarizes the significant data.

There were no significant achievement differences,either on the immediate achievement tests or on the

retention test, nor did the amount of time practicedby the groups differ.

The results of this study suggest that subjects whopracticed words on the microcomputer were less ableto assess their current state of knowledge than stu*dents using traditional techniques. Several alterna.*tive explanations are possible. For example, thesestudents may have focused less attention on their owncognitive activities. Alternatively, these studentsmay have perceived the computer as doing the learningfor them. In any case, further research on theeffects of CAI on metacognitive processes is warren*ted.

Table 1Summary of Data for Group Comparisons

Variable

Predictive InaccuracyNumber of sentence trialsNumber of definition

trials

Session P

Computer

A

T*Tests

Computer Traditional

2.2914 .92311.73 3.76

<.001<.001

4.9 3.0 <.05

Correlations

Traditional

r P A r

Day 1 3.46 1.46 .0654 2.0 .833 .757Day 2 4.33 2.0 .3128 2.2 2.1 .890Day 3 4.27 2.54 .4728 2.75 1.75 .683Day 4 4.47 2.43 .0245 3.38 3.23 .690

P = Predicted score (from total of 5)A = Actual score (from total of 5)

FUNDINGSOURCE

U. S. Department of Education, Office of SpecialEducation and Rehabilitative Services, Special Educa*tion Programs.

RESEARCHER Jacqueline HaynesFaculty Research AssociateInstitute for the Study of ExceptionalChildren and Youth

University of MarylandCollege Park, Maryland 20742301 ..454 *6921

112

HOFMEISTER

STUDY A Study of Selected Microcomputer-based ArtificialIntelligence Systems in Special Education

SUMMARY The academic and industrial worlds have shown exten-sive interest in knowledge engineering and the deve-lopment of expert systems. Expert systems promise tosupply software products that will increase theavailability of expertise in important areas. Inaddition, the associated product development activi-ties should also accelerate the development andclarification of the knowledge area itself.

Major reasons for the lack of artificial intelligence(AI) applications in education are the lack oftrained personnel and lack of necessary, but expen-sive and specialized, hardware and software. Theseproblems can be countered if recently developedsoftware can be shown to validly emulate some of theexpertise of special educators. These recently deve-loped software products have been designed to reducedependence on expensive computer programming ser-vices.

The research question is, "To what extent can selec-ted expert systems emulate the problem solving ofspecial educators?" The research is restricted todiagnostic activities and instructional prescriptionin mathematics and language arts.

This project should have implications for the workof:

1. Policy makers responsible for the allocation ofresearch resources in special education. The Secre-tary of the Department of Education has alreadycalled for the application of artificial intelligenceto the problems of education. Research on the valueof different AI applications is needed.

2. Developers interested in AI product development.Product development efforts will be enhanced if deve-lopers have a research base in specific AI applica-tions.

There is a need to create an acceptable R&D model forexpert systems in education. Project staff andothers are pursuing efforts in this area.

FUNDING U.S. Department of Education, Office of Special Edu-SOURCE cation and Rehabilitative Services/Special Education

Programs

113

RESEARCHER Alan HofmeisterDirector, Artificial Intelligence Research and

Development UnitProfessor of Special EducationUtah State UniversityUMC-68Logan, Utah 84322801-753-7973

S

HOWELLJURICA

STUDY The Effects of Computer Use on the Generalization ofLearning with a Learning Disabled Student

SUMMARY Statement of Problem. Although there are a number ofarticles coda-F.1717(i the use of microcomputers withspecial populations, there is little actual researchbeing done on the effectiveness or impact of the useof microcomputers with special populations(Hofmeister, 1982; Blaschke, 1985). This isespecially true as it concerns students who are diag-nosed as having learning disabilities (Schiffman,Tobin and Buchanan, 1982). It may be that the use ofcomputers and educational software will facilitatethe ability of this type of student to stay on taskfor longer periods of time and for the learning thattakes place to be generalized outside of the compu-ting situation. This study will attempt to investi-gate the effects of the use of computer and mathema-tical software on the attending skills and generali-zation of learning with a learning disabled studentin a special education setting.

Hypothesis 1: Students will increase the totalamount of time attending to mathematics-related tasksas a result of the use of computer-based mathematicsmaterials, as measured by an increased amount of timein non-computer-based mathematics activities.

Hypothesis 2: Students will generalize the informa-tion gained by using computer-based mathematicssoftware as measured by increases on a posttest ofmathematics achievement.

Independent Variables: Computer and mathematicssoftware (Plato: Basic Number Facts, 1982)

Dependent Variables: Performance measures on pretestand posttest of basic mathematics facts (KeyMathDiagnostic Arithmetic Test, 1976)

Procedures. The student selected for this study willbe randomly selected from among a class for thelearning disabled within the Columbus Public Schools.The treatment will consist of four periods in whichthe student will be observed. The first period willbe a baseline observational session, in which thestudent will be observed within the classroom as theywork on traditional mathematics assignments. Thesecond period will be a treatment condition, in which

115

FUNDINGSOURCE

RESEARCHER

the student will be exposed to the computer-basedmathematics materials for 10 minutes per day for 10days. The third period will be a return-to-baselinecondition in which the student is observed within theclassroom as the class works on traditional mathema-tics materials. The fourth will be a return to thetreatment period.

The data collection will involve both traditionalquantitative academic performance measures in theform of a pre- and posttest, and applied behavioralanalysis methods in the form of an observationalrecording technique.

Data Analysis. Data analysis will involve a t-testfor correlated means to analyze the resultant datafor the pre- and posttests. Data from the observa-tional recording sessions will be encoded and thendisplayed in the multiple-baseline format, as graphicrepresentations of the data.

The implications of this research are twofold:

1. The issue of generalization of computer-assistedlearning is of great importance to educatorsbecause of the increasing use of software inspecial education classes. If generalizationdoes occur in similar situations, then the com-puter may be eTiMiga as a valuable assistivedevice in the special education environment.

2. There is very little single-subject researchbeing done at this time in the area of computerapplications. It may be that this is a valuableresearch technique that will generate behavioraldescriptions and data which is not availablefrom group research studies.

The Ohio State University

Richard D. HowellAssistant ProfessorThe Ohio State UniversityCollege of Education225 Ramseyer Hall29 W. Woodruff AvenueColumbus, Ohio 43210-1177614-422-4872

and

116

James JuricaThe Ohio State UniversityCollege of Education225 Ramseyer Hall29 W. Woodruff AvenueColumbus, Ohio 43210-1177614-422-4872

117

STUDY

SUMMARY

FUNDINGSOURCE

RESEARCHER

HUMMEL

Special Purpose Software and Adaptive Peripherals: ACase Study

This is a case study of a 7<year<old boy in a firstgrade class in a special school. The boy is diag<nosed as having cerebral palsy. He is wheel chair<bound with limited use of his arms and hands and heis basically nonverbal, although he has some successwith speech sounds as a form of communication. Thisis an investigation of the application of micro<processor technology to this case. The study tracesthe progress from the parents' identification of thevarious features of the problem, to the provision ofadaptive hardware (including a specialized wheel<chair), and to the development of special purposesoftware and the integration of this hardware andsoftware into the child's educational program. Itfocuses on the efforts of the parents to secureeducational placements with the potential to appro<priately serve their son and to secure adaptive hardware. A morse code communication software programwas developed especially for their son, and it hasbeen integrated into his school program.

Parents and teachers have provided very favorablereports of the integration of the software and hard<ware. However, the parents have logged many hours topromote this progress. There have been somedifficulties getting full implementation in theclassroom.

The implications of this investigation are thatmicroprocessor technology can benefit a multiplyhandicapped young child, but only with substantialand sustained parental effort. Much more needs to bedone to facilitate access to technology, and thedevelopment and match of ameliorations involvingtechnology.

Progress in reading sight vocabulary has beenmeasured and results are being analyzed.

None Specified

Jeffrey W. HummelAssistant Professor209 Christopher Baldy HallState University of New YorkUniversity of BuffaloBuffalo, New York 14260716<636<2454

118

so

STUDY

MACARTHUR

An Examination of Learning Disabled Students' Composition in Three Modes: Handwriting, Word Processing,and Dictation

SUMMARY Writing -Is a complex process that involves a largenumber of skills at different cognitive levels.Writers must attend to handwriting, spelling, punctuation, usage, style, ideas, purpose, organization,audience, and many other factors. For skilledwriting to occur many of the lower level skills mustbe automatic. The higher level skills must be coordinated to permit conscious attention to be sharedamong them. Word processing has the potential tohelp LD students in several ways. The ability toprint a neat, error free copy can be tremendouslymotivating and positively affect students' selfevaluations. For those with handwriting problems,typing and the ease of making corrections may reducethe demands of producing text. The ease of revisionmay reduce concern with conventions enough to enablestudents to concentrate more on content and organization.

The purpose of this study was to describe the writingprocess and written products of learning disabledstudents using three different composition modes:handwriting, word processing, and dictation. Theresearch combined a case study approach with groupcomparisons among composition methods.

Eleven fifth and sixth grade learning disabled students experienced in using a word processor participated in this study. Background information on students' writing in the natural school setting wasgathered in several ways. First, all availablewriting done by the students in the current year wascollected from word processor data disks, journals,and writing folders. Second, special and regulareducation teachers were interviewed about theirwriting programs in general and the writing done byeach of the students. Third, the students wereinterviewed about their writing. Fourth, componentwriting skills were assessed through administrationof the Test of Written Language, measures of handwriting grid tyliiirged, and a measure of wordprocessing capability. Finally, writing instructionwas observed in students' classes.

IIIEach student wrote three stories, one with eachcomposition method. Students had two individual

sessions for each story. In the first session, thestudent was shown a picture and asked to write astory about it. In the second session, the studentswere asked to make any revisions they thought wouldimprove the story. Sessions were videotaped. Video-tapes were analyzed into time spent on prewriting,production, reviewing (re-reading), revision,pausing, and off-task behaviors. Spontaneous com-ments of the students and answers to probe questionsduring and after writing were used in interpretingthe planning activities of students. Revisionswithin the first draft and at the second session werealso analyzed.

Several quantitative measures were used in analyzingthe resulting stories including the following: num-ber of words, number and length of T-units, mechani-cal errors, grammatical errors, and vocabulary diver-sity. A holistic rating of overall quality and a

primary trait analysis focused on story grammar ele-ments were also used. In addition, studentsevaluated their own stories.

As the data have just been collected, results are notyet available. However, a final report will be com-pleted by May, 1985.

Word processing has considerable potential as awriting tool for LD students. However, research onits use is needed to clarify the benefits and prob-lems so that its potential can be realized. Inaddition, the first study contributes to the limitedliterature on the writing skills of learning disabledstudents. Previous research has focused only on thewritten products; whereas, this study examines thewriting process and revisions as well.

FUNDING U. S. Department of Education, Office of SpecialSOURCE Education and Rehabilitative Services/Special Educa-

tion Programs.

RESEARCHER Charles MacArthurFaculty Research AssociateUniversity of MarylandISECYBenjamin Building, Room 1220College Park, Maryland 20742703-454-5427

120

HURLEY

STUDY Publisher's Considerations Toward Software Researchand Development

SUMMARY Software publishers have become increasingly moresophisticated regarding the development of microcom-puter software for special populations over the pastseveral years. However, three major issues remain tobe addressed if the development of special educationsoftware (and other technologies) is to keep pacewith software development for nonhandicapped per-sons. The issues to be addressed include the follow-ing:

1. The Market. Is there a viable market in softwarefor speCTETgroups? Can publishers market enoughsoftware to recover development, manufacturing, pro-motion, and advertising cost? How does a publishereffectively tap the special education market? - OR -Where are we? Where are we going? Where doesspecial education fit into the Scheme of Things?

2. From Theory to Practice How can publishers keepabreast of The current research? How can publishersidentify and collaborate with researchers on an on-going basis? How should/can a publisher interpretbasic research to develop practical software foruaily use in the classroom? - OR - What does allbasic research really mean?

3. Collaboration. How can publishers collaboratewith: (a) researcher to ensure timely, technicallyadvanced, instructionally sound software? (b) DOE andprofessional organizations to ensure dissemination ofrelevant research information, joint funding of pro-jects, etc.? (c) key technology people to serve onsoftware that is educationally sound and appropriateto the needs of handicapped learners? (d) hardwaremanufacturers to plan low cost headphone jacks; voiceactivated units to ensure that as many modes oflearning as possible are utilized? (e) SEA and LEApersonnel to assist in field testing and crash test-ing to ensure that software products will be utilizedeffectively in the field? Finally, how can publishersgain seats on advisory committees of DOE and profes-sional organizations' technology committees to ensuremarketability of proposed products and disseminationof information to other publishers through the Soft-ware Publishers Association? - OR - How far can wego without a little help from our friends?

RESEARCHER Kathleen M. HurleyVice-President, Microcomputer SoftwareGrolier Electronic Publishing95 Madison AvenueNew York, New York 10016212-696-9750

KULIK

STUDY Synthesis of Research Findings on the Effectivenessof ComputerBased Education

SUMMARY The purpose of our research has been to answer basicquestions about the effectiveness of programs ofcomputerbased education (CBE) : How effective hasCBE been in improving instructional effectiveness?Has it been especially effective in bringing aboutcertain types of educational outcomes? Have certaintypes of CBE programs been more effective thanothers? Have certain types of evaluation designsbeen especially useful for showing the true effectsof CBE?

The method that we used to answer these questions wasmetaanal sis -- the quantitative analysis of resultsnisn.;a collection of studies for the purpose ofdrawing overall conclusions about the topic. Metaanalyses locate studies of an issue by objectivesearches of the literature; they express features andoutcomes of each study in quantitative terms; andthey statistically analyze the aggregate results.

RESULTS Computer searches of library data bases yielded 199separate, controlled evaluations of the effects ofCBE on learners. The studies covered elementaryschool, high school, and postsecondary applicationsof the computer in a variety of subject areas.Statistical analyses of the collected results showedthat CBE programs have generally had positiveeffects, as measured by several different criteria.These effects were not uniformly high, however, forall types of CBE programs at all instructionallevels. Nor were effects equally clear for everytype of research design. Study outcomes also differed somewhat in published and unpublished studies.

If future CBE programs are as carefully designed aspast programs have been, then they will most likelyproduce positive results. Development of CBE programs for the schools should therefore be encouraged.The design of such programs, however, should takeinto account evidence on the somewhat different effects of different types of CBE programs. In addition! research is needed on (a) the factors that haveproduced somewhat different results in published andunpublished evaluations of CBE, and (b) the factorsthat have produced somewhat different results instudies using different evaluation designs.

e

4)

FUNDINGSOURCE

RESEARCHER

National Science Foundation, Exxon Education Foundation

James A. KulikResearch ScientistCenter for Research on Learning and Teaching109 E. Madison StreetAnn Arbor, Michigan 48109313-764-0505

124

LAMOS

STUDY Microcomputer Software for Individually ManagedInstruction -- The H.E.L.P. Authoring System

SUMMARY This project involved the development of a microcom-puter-based authoring system for special education.The system developed provides two levels of author-ing capability in recognition of the different rolesthat special education and regular classroom teachersplay in achieving individualized instruction for thehandicapped child. At the first level, called Micro-Level Authoring, a teacher is able to produce small,modular units of instructional courseware that covera single concept and use a particular instructionalstrategy. These modular units of courseware arereferred to as micro-units. A micro-unit is intendedto involve the student in 10 to 15 minutes ofinstruction.

At the second level of authoring, called Macro-LevelAuthoring, a teacher can select and arrange specificmicro-units to form a lesson-unit. Lesson-units areused to meet specific curriculum objectives as wellas satisfy the requirements of a student's particularlearning need. At the Macro-Level, the teacherassigns one or more lesson-units to a student via astudent disk. Each student has his/her own studentdisk. The delivery program of the system "reads" thestudent disk and presents lesson-units and theircomponent micro-units to the student in the propersequence.

The production of the microcomputer-based materialswith this system does not require knowledge of acomputer language. The authoring system uses "edi-tors," containing menus and prompts, to elicit theinformation necessary to create instructional course-ware. Graphic presentations, including colored text,can also be created.

The authoring system has been developed in UCSD Pas-cal. In addition, the software incorporates theinternational GKS (Graphic Kernal System) graphicsstandards. The use of these standards in combinationwith the p-System operating system from Softech Inc.insures a high degree of software portability betweendifferent microcomputers. This means that instruc-tional materials created on one machine can be trans-ferred to another machine, without having to be re-written or recoiled. The authoring system ispresently running on the IBM PC and will be available

on the 128K Apple Ile and the Apple Macintosh in thefuture.

FUNDING U. S. Department of Education, Office of SpecialSOURCE Education and Rehabilitative Services/Special Education Programs

RESEARCHER Joseph P. LamosSystem AnalystDenver Research InstituteUniversity of DenverDenver, Colorado 80210303-871-2420

STUDY

MCGREGOR

Applications of Technology in the Education ofSeverely and Profoundly Impaired Students: Research,Training, and Information Dissemination.

SUMMARY The overall goal of this 2<year project is toincrease the utilization of technology in classroomsserving severely and profoundly impaired students by(1) identifying those aspects of classroom managementthat can be improved or made more efficient throughthe use of personal computers and related softwareprograms; (2) investigating the effects of computerutilization on the quality of instruction provided tothe students and their performances relative to theirgoals and objectives stated on their individual edu<cation program; (3) providing training to specialeducation supervisors, instructional advisers, andteachers to extend the techniques developed throughthis project to other schools and classrooms through<out Philadelphia; and (4) disseminating informationabout successful project activities to interestedprofessionals on a national basis. The project willbe a collaborative effort of the Division of Educe<tion of the Johns Hopkins University and the Divisionof Special Education of the School District of Phila<delphia.

The 12 classrooms participating in the project arelocated in schools that serve both handicapped andnonhandicapped students and were selected to achievediversity in student ages and handicapping condi<tions. Four of the rooms are in high schools, fourare in junior high or middle schools, and four are inelementary schools. During the first year, six class<rooms (two in high schools, two in junior high ormiddle schools, and two in elementary schools) willbe randomly selected to participate in a treatmentcondition, with the remaining six classrooms servingin a control condition. In the second year, theproject will be extended to six additional classroomsproviding both within< and between<classroom compari<sons of results.

Each classroom will be equipped with an Apple liesystem (computer, monitor, and printer) and relatedhardware and software items. Five of the classroomswill also receive Texas Instrument 99/4A color compu<ters equipped with expansion boxes, monitors, andLogo II. The hardware and software will be used (1)to facilitate the planning and design of effectiveinstructional programs; (2) for instructional pun*

127

poses, particularly in the domain of interpersonalcommunication and interaction; and (3) to facilitatethe collection, analysis, and utilization of datathat measure the performance of each student relativeto his/her educational goals and objectives. Theprimary measure of the efficacy of employing microcomputers for these purposes will be observationaldata obtained at regular intervals reflecting thetypes of tasks and activities engaged in by teachersand students.

The participating classrooms will serve as sites fortraining special education supervisors, instructionaladvisers, and teachers in the hardware and softwareapplications developed and/or assessed by the project. A number of audiovisual presentations and amanual describing these applications will be developed during the latter part of the second projectyear for distribution in the Philadelphia schools andto interested professionals throughout the country.

FUNDING U. S. Department of Education, Office of SpecialSOURCE Education and Rehabilitative Services/Special Educa

tion Programs.

RESEARCHER Gail McGregorProject CoordinatorStevens Administrative CenterRoom 30413th and Spring Garden StreetPhiladelphia, Pennsylvania 19123215-351-7007SpecialNet: IU26

MALOUF

STUDY Research Project on the Effectiveness of Microcomputers in Special Education

SUMMARY The two broad areas being investigated in this project are the effectiveness of microcomputers inimproving the learning of handicapped students andthe contextual factors associated with these benefits. The research is being conducted with mildlyhandicapped elementary and middle school students.Current and planned studies focus on instructionalgames, metacognition, word processing, behavioral andaffective outcomes, response latency, software evaluation, student groups and learning strategies.(Specific studies are described in the research summaries of C. MacArthur and J. Haynes.) The majorimplications are that research efforts must continueand expand.

Issues encountered in conducting this research include:

1. Defining research questions in their most usefulform.

2. Selecting specific studies from so many alternatives.

3. Designing research software and measurement instruments.

FUNDING U.S. Department of Education, Office of Special EduSOURCE cation and Rehabilitative Services/Special EducationPrograms

RESEARCHER David MaloufProject DirectorInstitute for the Study of Exceptional Children andYouth

University of MarylandDepartment of Special EducationBenjamin Building, Room 1220College Park, Maryland 20742301-454-6921

129

a

STUDY

MOLL

Access Knowledge Base (Knowledge Base on ComputerAccess and Microcomputer-Based Aids for Persons withPhysical and Sensory Handicaps)

SUMMARY Development of the Access Knowledge Base has twoprimary goals: (1) to provide information that couldenable an inexperienced user of the system to findpotential solutions to problems related to computeraccess; and (2) to promote communication and collab-oration between current or potential computer opera-tors with physical or sensory handicaps, researchers,developers of aids and equipment, rehabilitationprofessionals, and employers in the knowledge indus-try.

In addition to descriptions of commercially availablemicrocomputer-based aids and products facilitatingcomputer access, the knowledge base is expected toprovide lists of other relevant information resourcessuch as periodicals, directories and guides,clearinghouses, and computerized networks and data-bases. It will also list computer manuals availablein large or nonprint media. Finally, it will supplyinformation about centers that provide evaluation orother services related to computer access; companiesor agencies that sponsor training programs for dis-abled persons who want careers requiring the use ofcomputers; and centers performing research in therehabilitation area.

For the most part, the system will be menu-based tofacilitate single switch input. It will also bedesigned to encourage users to pose questions and tomake comments on specific topics. These will beautomatically routed to designated content experts inthose areas. Through opening communication channelsin this way, the available store of informationshould be expanded. Further, such forms of communi-cation would hopefully enhance the ability of resear-chers and developers of aids and computer interfacesto be more responsive to the needs of consumers.

It is probable that the knowledge base will not beavailable fox access to the general public until thefall or winter of 1985. We are not yet certain howthe system will be accessed.

FUNDING None specifiedSOURCE

130

RESEARCHER Judy MollContent Coordinator for Knowledge Base on HandicappedAccess .

Department of Medical Information ScienceMicrocomputer Systems Laboratory1408 West University AvenueUrbana, Illinois 61801217-333-9277

131

STUDY

SUMMARY

FUNDINGSOURCE

RESEARCHER

MOLL

Summary of Research on Executive Workstations forPersons With Physical Handicaps

The objective of this research is to improve theproductivity of persons with physical handicaps asthey pursue an academic or managerial career. It isbelieved that the productivity of handicapped personscan be enhanced through applications of technology toroutine office procedures which are handled quicklyand effortlessly by nonhandicapped persons.

The UNIX operating system environment has been selected as a focus for research into aspects of an operating system that can be modified to meet the specialneeds of handicapped users. The choice of the UNIXenvironment will assure the portability of softwaredeveloped for the project to new systems that supportthis environment.

It is expected that longterm research will includeexamination of technological developments in thefollowing areas: digitized voice recording fromphone conversations; recorded voice output messagesfor support of computerassisted conversation (byspeech impaired persons); automated input fromprinted material, with storage of graphic images fordiagrams; online reference material, such as dictionaries or scientific reference tables; flexible,interactive input (from standard keyboard to singleswitch or limited audio input); online phone directory and dialing (for regular conversations as wellas for communication with computer systems); andcomplete paper transport output system (for documents, graphics, and diagrams).

None specified

Judy MollContent Coordinator for Knowledge Base on HandicappedAccess

Department of Medical Information ScienceMicrocomputer Systems Laboratory1408 West University AvenueUrbana, Illinois 61801217-333-9277

132

STUDY

SUMMARY

e

FUNDINGSOURCE

RESEARCHER

MOORELAHM

Robotics, Artificial Intelligence, Computer Simula-tion: Future Uses in Special Education

This presentation reports on a study which had twoprimary objectives:

1. To enumerate and describe the ways in whichrobotics, artificial intelligence, and computersimulation are being applied in other fields; and

2. To identify potential uses of these technologiesfor special education students.

The study was conducted using a new methodology,developed to identify existing applications oftechnologies in one field and forecast their futureapplicability to situations in other fields. Thefive steps of the methodology are discussed in thepresentation.

The presentation briefly defines the three technolo-gies, and describes the "state-of-the-art" of eachone, as evidenced by their uses in settings otherthan education. The presentation also: (1) setsforth the conclusions from the study about the threetechnologies; (2) enumerates the general conclusions,based on ratings by the expert panelists of scenariosof possible future uses of the technologies; and (3)provides specific conclusions related to individualscenarios.

The need for a new methodology for studying theadvanced technologies was identified during thestudy. A new method was developed and is brieflydescribed in the presentation.

U.S. Department of Education, Office of Special Edu-cation and Rehabilitative Services/Special EducationPrograms

Gwendolyn B. MooreProject DirectorThe COSMOS Corporation1735 Eye Street, N.W.Washington, D.C. 20006202-728-3939

and

133

Elizabeth A. LahmResearch AssociateThe COSMOS Corporation1735 Eye Street, N.W.Washington, D.C. 20006202-728-3939

134

MOROCCONEUMAN

STUDY Microcomputers, Writing, and Learning DisabledChildren: Field-based Research

SUMMARY Writing problems are the most prevalent communicationdisability of learning disabled (LD) children. Thisstudy focuses on the role of word processing programs(text-editors) in teaching writing, since severalfeatures of text-editors make them promising writingtools for LD children. Four objectives of this 2-year, classroom-based study are to:

1. Determine whether LD children can develop text-editor facility;

2. Identify the writing strengths and problems ofLD children and identify the text-editorfeatures that can help them build on thosestrengths and improve their writing skills;

3. Explore how remedial teachers can integratetext-editors into their teaching of writing, howstudents respond to those approaches, and howthey can improve LD children's writing;

4. Develop training materials that enable remedialteachers to use text-editors to improve LDchildren's writing skills.

In the first six months we documented the diverseways five teachers taught writing to fourth grade LDstudents, using text-editors. Weekly observation,transcription of verbal interaction around the compu-ter, and interviewing techniques were used. It wasfound that for children to realize the benefits oftext-editors they need systematic training in typingand text-editing, and remedial teachers need furtherguidance in the teaching of writing. Therefore, wedevoted the second six months of Year One not only tocontinuing observation and analysis of classroomactivities, but also to providing teachers with feed-back and guidance and to developing a framework forteaching writing to LD students that is based oncurrent advances in writing research.

No studies of LD children writing with text-editorsare based on the children having received extensivetraining in typing and text-editing. Continuingresearch should include this element, in order todiscover the maximal possibilities of this computertool for LD children.

The perspectives on writing and the assumptions aboutLD writers which many remedial teachers bring to theuse of texteditors may not lead to productivewriting experiences for LD children. This suggeststhat special training may be required in order forremedial teachers to maximize the potential benefitsof word processing for LD children.

RESULTS A major finding of Year One is that remedial teachersbring three different perspectives on writing totheir use of texteditors, and the approaches differsubstantially in the extent to which they actuallyuse the texteditor to foster a sense of writingownership in the LD child. Year Two will focus ondeveloping and fieldtesting classroom approaches andactivities based on the framework and findings ofYear One.

FUNDING U. S. Department of Education, Office of SpecialSOURCE Education and Rehabilitative Services/Special Educa

tion Programs.

RESEARCHER Catherine Cobb MoroccoSenior Research AssociateDirector, The Writing ProjectEducation Development Center, Inc.55 Chapel StreetNewton, Massachusetts 02161617-969-7100 ext. 270

and

Susan NeumanThe Writing ProjectEducation Development Center, Isic.55 Chapel StreetNewton, Massachusetts 02161617-969-7100

136

STUDY Technological Needs and Applications

SUMMARY

FUNDINGSOURCE

RESEARCHER

MORRIS

Administer programs targeted at materials developmentor adaptation for use by blind students. Needs wereidentified and prioritized by meetings of consultantswho are knowledgeable and active in using microcompu-ters in educational applications with blind students.Additional input was received by means of a nationaltechnology/microcomputer questionnaire used to deter-mine perceived needs and to determine the "state-of-the-art" for educational applications with blindstudents.

The American Printing House for the Blind is thiscountry's primary source for special educationalmaterials to meet the unique needs of visually handi-capped students. The results of the research anddevelopment program will be products blind studentscan use so they can benefit, like their sightedpeers, from use of microcomputers in the classroom.These products will be distributed by the AmericanPrinting House for the Blind.

Projects resulting include: preparation of brailleeditions of a literacy book and instructionalmanuals, development of a Basic Familiarity Program,modification of the Textalker Speech Generator (EchoII) spelling program and list builder, and contactswith software publishers regarding permission toadapt programs to make them speak. Evaluations areon the order of legibility studies and in-usecritiques by teachers using new programs.

American Printing House for the Blind

June E. MorrisDirectorDepartment of Educational ResearchAmerican Printing House for the Blind, Inc.183.? Frankfort AvenuePost Office Box 6085Louisville, Kentucky 40206502-895-2405

NAIMAN

STUDY Telecommunications and an Interactive ProcessApproach to Literacy in Deaf Young Adults

SUMMARY This research was designed to seek solutions to acentral problem in the education of deaf students,their severely limited ability to read and writeEnglish. The research developed a conceptual modelthat combined and brought to bear on this problem twomajor areas of recent development in technology andtheory. The project model utilized electronic communicati.m in a way that is consistent with currenttheory and practice in literacy development as aninteractive process. The model was implemented,fieldtested, and evaluated in five schools for thedeaf.

The research was supported by a DistinguishedProfessor Award from the National Institute forHandicapped Research. The General Telephone andElectronics Corporation (GTE) provided the telecommunication equipment needed for the project. Equipment and telecommunication services included installation of Electronic Mailboxes for all of thestudents in the project; establishment of anElectronic Bulletin Board, The Deaf StudentsNewswire; establishment of a Computer Teleconferencing System that enabled the researcher, the AdvisoryCommittee specialists, and the cooperating teachersto have ongoing dialogue with each other; and provision of a special computer program and facilities foranalyzing the vast amount of written communicationgenerated in the project.

Five schools for the deaf participated. Each schoolprovided a cooperating teacher and six students.Cooperating teachers received ongoing assistance fromthe project director on theory and methods of providing a functional reading and writing environment tofoster maximum growth in literacy.

A group of distinguished deaf and hearing specialistsin literacy acquisition and telecommunications servedon an Advisory Committee for the project and providedlinkages with the Center for Applied Linguistics atGallaudet College, the Communication Research Centerat National Technical Institute for the Deaf, and thePrinceton Research Forum.

Evaluation of the project included both summativeevaluation of students' growth in reading and writing

FUNDINGSOURCE

RESEARCHER

skills and formative evaluation analysis of thevarious project components.

U.S. Department of Education, National Institute forHandicapped Research

Doris W. NaimanProfessor of Deafness RehabilitationNew York UniversityDivision of EducationDepartment of Communication Arts and Science829 Shimkin HallWashington SquareNew York, New York 10003212-598-2921, 2922, 2923

139

STUDY

O'DONNELL

The Effect of Microcomputer Training in LOGO and WordProcessing on Achievement of Children with LearningDisabilities

SUMMARY The proposed research will investigate the hypothesisthat microcomputer training in LOGO (a nonverbal,spatial program) and/or word processing (a writtenlanguage program) results in increased academicachievement for children who have learning disabili-ties.

A pretest-posttest control group design is planned.A total of 50 students between the ages of 6 and 12years, who are eligible for special education ser-vices for learning disabilities, will be randomlyassigned to two groups. Students will be given apretest as part of the in-depth 8-week assessment atthe UMKC Learning Clinic. These students' learning-- capacity, achievement in academic skills, andintra-individual differences -- will be assessed. Atthe end of the assessment period, Group I will begina 30-week microcomputer training during the child'sregular school day in a school district in the KansasCity area. The microcomputer training will consistof daily sessions with graduate research assistants.Group II will receive no treatment; however, ateacher-aide will provide Group II students with anactivity for the same period of time and to the samenumber of children as Group I. In the final twoweeks of the proposed project, the posttesting willbe made and a double-blind comparison of pretest andposttest data from Groups I and II will follow.

The proposed study is seen as a feasibility study, asa beginning phase for a series of studies on theeffect of the computer when used with learningdisabled students in assessment and remediation.Four research questions will be considered. Thefirst is experimental and will be analyzed withfactorial analysis of variance methods. The secondis exploratory and will be evaluated by means ofcorrelational methods. The third and fourth requireclinical observation and will be addressed throughdescriptive analysis.

1. Primary research question. Will microcomputertraining in LOGO and/or word processing increaseacademic achievement for students who have severeor moderate learning disabilities?

140

FUNDINGSOURCE

RESEARCHER

2. Primary research question. Is there a relationship between the content area of the disability(e.g., reading, spelling, math, writing, language) and the ability to acquire LOGO and/orword processing?

3. Secondary research question. Can severely andmoderately -3 nFETrig disabled students at theelementary school level learn microcomputerprograms in LOGO and/or word processing?

4. Secondary research question. What adaptationsare requi70-737 disab/ed students to useexisting LOGO and word processor programsdesigned for young children? Is there a need fora special training program for learning disabledstudents in microcomputer uses?

The results of the study may have implications forsoftware development and /or instructional methods forteaching computer programs to learning disabledstudents. In addition, it is expected that theinteraction between learning difficulties and thecomputer programs for LOGO and word processing willbe better understood.

Comparative instructional features of the numerousword processing programs has been a major issue indesigning and conducting this technologybasedresearch.

Weldon Spring Endowment Fund, University of MissouriKansas City

Linda Elizabeth O'DonnellAssociate Professor and Special Education

Program Coordinator365 Education BuildingUniversity of MissouriKansas City5100 Rockhill RoadKansas City, Kansas 64110816-276-2254SpecialNet: MOUMKC

141

PANYANHAZAN

STUDY An Authoring System for Individualized Instructionand Applied Research

SUMMARY The purpose of this program was to develop and fieldtest a prototype microcomputerbased multisensoryauthoring software system for special education,designed to improve the individualized instruction oflearning disabled children.

An overview and description of a new authoring systemfor special education will be presented. Initialfieldtest results for six LEAs will be shared.

The authoring system as a vehicle for research aswell as teaching will be discussed. First, the system affords a comparison of an adaptive learningmodel (through program branching and teacher generated changes) with a program without these features.Secondly, the use of the authoring system to examineprocess and content variables on learning rates andpatterns will be illustrated.

The proposed model for researching learning variablesis "userfriendly" and has direct implications forinstruction and success during instruction. Alsosince data are collected by the computer program,teacher acceptance should be maximized.

One issue encountered during this research investigation was the difficulty of obtaining a consistentindependent variable when the nature of the intervention is changing due to necessary adaptations andadjustments.

FUNDING U. S. Department of Education, Office of SpecialSOURCE Education and Rehabilitative Services/Special Educa

tion Programs

RESEARCHER Marion V. PanyanAssociate Professor of EducationThe Johns Hopkins UniversityDivision of Education100 Whitehead Hall3400 North Charles StreetBaltimore, Maryland 21218301-338-8273SpecialNet: JHUL

and

142

Paul L. HazanAssistant to the Director for AdvancedComputer Technology

The Johns Hopkins UniversityApplied Physics LaboratoryJohns Hopkins RoadLaurel, Maryland 20707301-953-5000 ext. 7449

143

STUDY

PRINZ

ALPHA MicrocomputerVideodisc Project for YoungChildren: An Interactive Approach to Reading, Computer Writing and Communicative Development

SUMMARY This research project involves a new method forteaching young children to read and "write." Thestudy incorporates the use of a portable microcomputer and videodisc player in the classroom to allowthe child to initiate communication from the veryonset of instruction. Traditionally, children usingcomputers and interactive videoassisted learninghave been required to respond to statements and questions and their responses have been required to fallwithin a small range of acceptable answers. Such anapproach provides the child with limited opportunities to initiate language or to be an active catalystin communication. The subjects included in the firststudy are 75 deaf and 25 multiply handicapped children between the ages of 2 and 15 years, with averageor better intelligence. The children are enrolled inschools in the Philadelphia and Pittsburgh metropolitan areas. The children are trained to use a novelinteractive microcomputer system with a specialinterface keyboard which builds in perceptualsalience, individualized vocabulary, animation ofpictures, and graphic representations of signs fromAmerican Sign Language (ASL). The child is taught topress keys with pictures, words, and short statementsdrawn from the child's own central interests andfavorite expressions. This is possible because (1)the keys are readily changeable; (2) new words andgraphic representations of pictures can be quicklyand efficiently entered in the computer; and (3)permanent disc storage allows instantaneous access tomany printed words, accompanying color graphics, andactual pictures from the videodisc.

RESULTS Children read sentences and write sentences and alsocommunicate to the teacher about the messages printed, pictured, and signed on the computer's displayscreen. The learning mechanism underlying this novelinstructional system is best characterized as responsive, interactional, and exploratory, reflective ofthe way in which most children acquire a first language. Preliminary results of the study have demonstrated a significant improvement in word and phraseidentification, reading comprehension, and basicsentence construction or writing. The increase inwriting, reading, and communication skills is attributed to exploratory learning -- not solely programmed

instruction -- which allows the child to flexiblyinvestigate the representation of various printedforms which relate to the child's own primary mode ofcommunication.

FUNDING U. S. Department of Education, Office of SpecialSOURCE Education and Rehabilitative Services/Special Educa

tion Programs.

RESEARCHER Philip M. PrinzAssistant professor, CMDISDivision of Special Education and Communication

DisordersThe Pennsylvania State University217 Moore BuildingUniversity Park, Pennsylvania 16802814-863-2018

STUDY

RIETH

An Analysis of the Instructional and ContextualVariables That Influence the Efficacy of Computer-based Instruction for Mildly Handicapped SecondarySchool Students

SUMMARY This project is a 4-year study investigating andreporting on the use and efficacy of microcomputer-based instruction with mildly handicapped secondaryschool students. Studies are underway using quanti-tative and qualitative methodology to examine theinstructional, contextual, and interpersonalvariables that may influence the effectiveness ofmicrocomputer-based instruction.

To date observational data have been collected in 53classrooms across three school corporations. Datahave been collected to analyze the impact of themicrocomputer upon curriculum content, teacher beha-vior, student behavior, instruction, and theteacher's instructional focus. In addition the dataare being analyzed to determine the impact of micro-computer-based instruction on student academiclearning time and achievement.

A comparison of special education classes where acomputer was in use versus classes where a computerwas not in use indicated: (1) very small differencesin teacher and student behavior, instruction, and theteacher's instructional focus, indicating in the ob-served classrooms that the computer had little impactupon the general ecology of the secondary classroom;(2) consistent with the greater prevalence of mathsoftware, computers were most likely to be used formathematics instruction; (3) students enrolled inclasses where the teachers used computer-based in-struction spent 24.1% of their time working with thecomputer; (4) paper/pencil activities were higher incomputer-use classroom than in noncomputer classessuggesting that the time created by using the compu-ter as instructor was invested in paper/pencil activ-ity which was found by Rieth and Frick (1983) toincrease the likelihood of student off-task behaviorby seven times; (5) computer class students were morelikely to be actively engaged than their counterpartsin noncomputer classrooms.

Interview data were collected from 51 teachers whoparticipated in the observational study and from 25randomly selected mildly handicapped adolescents.Key findings from the teacher interviews indicate

that: (1) teachers presently use computers predomi-nantly for mathematics instruction although theydesire to use them in a much broader range of subjectareas; (2) machines are used predominantly for drilland practice, to reward student performance orbehavior, or to present games that have no instruc-tional content; and (3) major obstacles to computeruse are the unavailability and inappropriateness ofcomputer software, location outside the classroom,and the potential misuse or overdependence on tech-nology; (4) characteristics of good software includedease of use or comprehensibility of directions, mini-mal reading required, appearance (sound, graphics,animation), and fun for studerts.

RESULTS Students interviewed were generally pleased withcomputer-based instruction and would like to use thecomputers more than they do currently because theyfelt that the computers helped them learn schoolsubjects. Fun was the most important softwarecriteria for students.

Presently, intervention studies are being conductedto determine the effects of the microcomputer oninstruction, the classroom ecology, student behavior,and teacher behavior. Issues being explored include:(1) the additive effect of the computer to pedogogi-cally sound instruction; (2) the effects of time oncompuer and software quality on teacher and studentbehavior; (3) the effects of different types andtiming of feedback on student performance; (4) theeffects of increased direct and computer-based in-struction upon academic engagement, achievement,grades, and attendance; and (5) the effects of dif-ferent training strategies and content upon the useof computers as teaching tools.

FUNDING U. S. Department of Education, Office of SpecialSOURCE Education and Rehabilitative Services/Special Educa-tion Programs

RESEARCHER Herbert RiethProfessor and DirectorCenter for Innovation in Teaching the Handicapped

(CITH)Indiana University2805 E. 10thBloomington, Indiana 47401812-335-5849SpecialNet: IUCith

ROMANCZYK

STUDY Direct Computer Instruction: Issues of Assessment,Effecciveness, and Generalization

SUMMARY Past research has not clearly demonstrated the rela-tive effectiveness of direct computer-assistedinstruction versus teacher-assisted instruction. Aseries of research projects were designed to addressthis issue as well as related issues such as appro-priate assessment procedures to select children whowould benefit from computer-assisted instruction,collateral behavioral management (self-stimulation,acting out, inattention), and generalization ofskills acquired with the computer to more typicallearning situations. Our subject populations havefocused on somewhat diverse groups including thoselabeled autistic, severely emotionally disturbed, andlearning disabled.

Two studies involved direct comparison of teacherinstruction with computer instruction. The firststudy was conducted with children labeled autistic,neurologically impaired, emotionally disturbed andmentally retarded. A second study was conducted withchildren labeled as learning disabled. Both studiesemployed a similar methodology in that all subjectsparticipated in both conditions. In one condition anApple computer was programmed to deliver instructionin mathematics. In the parallel condition (matchedfor pacing, difficulty level, degree of interchange,reinforcement density, etc.) a teacher delivered asimilar instructional program. High quality voicesynthesis was utilized with the severely impairedgroup of children during computer instruction. Inthe study with the learning disabled children,psychophysiological measurements were taken to assessthe construct of attention in a multidimensionalfashion, that is, in addition to latency and perfor-mance measures.

A third study was conducted to assess the degree ofgeneralization of skills learned on the computer.This group of children was labeled as emotionallydisturbed and autistic and had basic pre-academicskills. Mathematics again served as the task andinstruction was delivered only via computer.Generalization measures were taken with respect toperformance on standard worksheets and flash cardtasks in the classroom setting.

The three studies also attempted to assess predictive

148

variables for child performance under computerinstruction and teacher instruction. A specificcomputer programming algorithm was developed thatmeasures a multifaceted array of child performancewhen using the computer. This includes not onlyrecording correct and incorrect responses, but alsopresence of no response, latency for each response,the degree of self-stimulatory behavior as indexed byrandom pressing of keys, number of type and repeti-tion errors and full pattern, and trial-by-trialanalysis. In addition, the software utilized allowedthe instructor to completely individualize allresponse-dependent parameters, that is, the type ofproblems, their degree of difficulty, their sequenc-ing, reinforcement density, reinforcement type, speedof presentation of problems, presence of verbal feed-back, visual feedback, negative feedback on incorrectresponses and positive feedback on correct responses.In this way, software can be tailored to the particu-lar idiosyncratic style of each learner. This soft-ware series now encompasses mathematics training withremedial tutorial exercises keyed to automaticallyengage based upon a series cf incorrect responses,matching-to-sample tasks, acquisition of alphabetskills, and spelling skills.

This line of research was designed to pursue the useof the computer as an assessment/diagnostic tool tostrip away social learning history from the specificsof the task that is to be taught. The effect uponcollateral behavior has been very strong in that withsome children there has been virtually a completesuppression of self-stimulatory and acting outbehavior with computer instruction, but not withteacher instruction. However, here again, there areindividual differences and this may allow us toanalyze in a more sophisticated fashion than haspreviously been possible the various motivatingfactors for "psychotic" behavior that in the pasthave been confounded with significant socialinteraction variables.

RESULTS The basic finding of our research has been thatcomputer-assisted instruction can be as effective asteacher instruction, but that it is both hardwaredependent (that is, it is very dependent upon theparticular sophistication of the computer system) andis somewhat population specific, although individualdifferences are apparent in all populations. We dohave a predictive model based on our results thatallows for subject selection as to who has thegreater probability of benefiting from such computer-assisted instruction. It appears that some of the

149

FUNDINGSOURCE

RESEARCHER

initial positive effects with learning disabledchildren are quite transitory and are tied to noveltyeffects, whereas with more severely disturbedchildren, the effects are more directly related tothe specifics of the instructional paradigm and thedegree to which the computer can carry these out in aflawless fashion. Generalization was easily achievedfor the majority of subjects, although this was notuniversally the case. Specifically, in one casestudy, there was no generalization until clinicalintervention for the child's nonresponding in theclassroom setting. Given this child's early historyof physical abuse and fear of interaction withadults, this particular paradigm has potential forassessment procedures above and beyond analysis ofacademic skill acquisition.

Children's Unit for Treatment and Evaluation, SUNYBinghamton

Raymond G. RomanczykAssociate DirectorPsychology DepartmentSUNYBinghamtonBinghamton, New York 13901607-798-2829

STUDY

SUMMARY

FUNDINGSOURCE

RESEARCHER

ROSEGRANT

Studies on the Use of Microcomputers to FacilitateLiteracy Development

Research was conducted to construct and test a modelof how the microcomputer can be used to foster pro<gress in reading and writing with children who haveexperienced continued failure in acquiring basicliteracy.

The methodology involved observations, analysis oftranscripts from videotapes, pre/postreading andwriting tasks, analysis of weekly writing by stu<dents, behavioral and perceptual measures. Most ofthe primary data was collected during children'swriting sessions using a microcomputer, a word pro<cessing program, and synthesized speech.

The implications of the research are:

1. The microcomputer can provide children with acontext for literacy learning through the use ofmultiple modalities.

2. The microcomputer can provide children withneeded forms of assistance which lead toincreased compentence in reading and writing.

The findings confirm that the microcomputer caneffectively provide forms of support, feedback, andrepetition, which enable children to make progress inreading and writing.

IBM Corporation, Apple Foundation

Theresa RosegrantAssistant ProfessorState University of New YorkUniversity of Buffalo593 Christopher Baldy HallBuffalo, New York 14260716<636<2453

151

STUDY

SUMMARY

S

RUSSELL

Microcomputers in Special Education: Beyond Drilland Practice

The overall project focuses on training, research,and dissemination of information about the use ofnon-drill-and-practice software with learning dis-abled and emotionally disturbed students in grades K-8. Objectives of the project include the developmentof training opportunities in cooperation with LesleyCollege Graduate School, the formation of a SpecialInterest Group for educators, and the production of ahandbook. The project also has two research com-ponents -- a national survey of current practice anda series of small classroom-based investigations.These two components are described below.

1. National Survey of Current Practice. A randomsample of 50 U.S. school districts, selectedfrom all school districts using computers in anyway for instructional purposes, was surveyed.Each district's special education administratorwas interviewed by phone. A teacher from eachof the districts in which computers were beingused for instructional purposes with specialeducation students was also interviewed. Thefocus population of the study was learning dis-abled and emotionally disturbed students ingrades K-8. One of the major research questionswas: What are the barriers to implementation ofbeyond-drill-and-practice uses of the computerwith this population? Data collection has beencompleted and data is currently being analyzed.A second phase of the survey, now underway, isthe identification of promising practices.Identification of such practices is taking placethrough press releases, announcements injournals and at conferences, and through profes-sional contacts. A central question in thispart of the study is: How have the barriersidentified in phase one been overcome by thosepractitioners who are successfully implementingsuch uses of the computer with their specialeducation students?

2. Classroom-based Investigations. These investi-gations are conducted as teacher-researchercollaborations. The first year of the projecthas included observations and case studies ofstudents using beyond-drill-and-practice soft-ware (Logo, word processing). Plans for the

152

secondyear include more detailed studies ofparticular uses of the computer. These studieswill be planned with teachers during the springand summer of 1985. We are particularlyinterested in the importance of teachers ascollaborators in research.

RESULTS Preliminary analysis of the survey results indicatethat the most significant implications of this dataare likely to be in the areas of teacher training andpolicy. There appears to be a poor match between theneeds of special education teachers and their students and the type of computer education which mostof these teachers have received. Poor articulationof goals or narrow interpretation of learning goalsalso appear to impede promising uses of the computerwith special education students.

Implications from the classroom investigations willbe practical ones -- what kinds of experiences workwell with particular students, what kind of teacherintervention is needed, how can outcomes be evaluatedby educators.

FUNDING U.S. Department of Education, Office of Special EduSOURCE cation and Rehabilitative Services/Special Education

411 Programs

S

RESEARCHER Susan Jo RussellProject DirectorTechnical Education Research Centers1696 Massachusetts AvenueCambridge, MA 02138617-547-0430

153

STUDY

SUMMARY

FUNDINGSOURCE

RESEARCHER

SAPONA

Writing Productivity of Learning Disabled Students:The Effects of Using a Word Processing Program

The purpose of this study was to begin the empiricalinvestigation of the effect of using a word proces-sing program on the writing skills of learning dis-abled elementary students. The following researchquestion was examined: Do learning disabled studentswho use a word processing program (Bank Street Wri-ter, 1985 version) to complete writing assignmentsdiffer on selected measures of writing productivityfrom students who complete writing assignments usinga pencil and paper?

Eighteen students in two learning disabilities self-contained classrooms served as subjects in thisstudy. All students wrote two stories per week for 8weeks. Students in the experimental group used aword processing program for writing during the last 4weeks of the study. Productivity measures includednumber of: words, different action verbs, differentdescribing words, different sentence beginnings,spelling errors, capitalization and punctuationerrors, and a measure of vocabulary diversity. Thedata are being coded and analyzed at this time.

This study was an initial investigation in what willlikely be a series of studies designed to determinethe most effective and appropriate instructionalinterventions to accompany the use of the microcompu-ter as a writing tool.

One outcome of this study was a set of files (activi-ties) for teachers to use in training elementary-agelearning disabled students to use the Bank StreetWriter program. Future studies will focus on thedevelopment of strategies of instruction as well asmore specific investigation of the writing process(e.g., revision strategies used by students who havedifficulty writing).

University of Virginia, Dissertation Research Award,Department of Special Education

Regina H. SaponaGraduate InstructorDepartment of Special EducationRoom 152 Ruffner Hall405 Emmet StreetUniversity of VirginiaCharlottesville, Virginia 22903804-924-7461

154

SEELEY

STUDY Research and Development of Microcomputer CoursewareAuthoring System for Learning Disabled Students

SUMMARY: Research Questions:

1. What are appropriate roles for regular andspecial educators in implementing new technologies at middle school level?

2. What is the appropriate role training forteachers to implement new technologies?

3. How can a microcomputer authoring system be bestimplemented with special and regular educationteachers at middle school level?

Methodology:

1. Observation2. Interviews3. Role Training4. Training in Instructional Design5. Implementation of Authoring System6. Evaluation7. Dissemination

FUNDING U.S.Department of Education, Office of Special EducaSOURCE tion and Rehabilitative Services/Special Education

Programs

RESEARCHER Ken SeeleyProject DirectorSchool of EducationUniversity of DenverDenver, Colorado 80208303-871-2508

STUDY

SUMMARY

SEMMEL

Project TEECh (Technology Effectiveness with Excep-tional Children)

Project TEECh is a federally funded, 4-year programof research on the effects and effectiveness ofmicrocomputer instruction with mildly handicappedchildren in elementary schools.

The investigators have developed a model to assessthe contextual variables which affect microcomputerallocations, access, and use patterns for mildlyhandicapped students. The impact of these contextualvariables is traced through a path model, from thosefactors most distal to microcomputer use in theclassroom (e.g., federal and state educational poli-cies) to those more proximal to actual Ilse (e.g.,teacher training and special education resources).Through this path model, the range of possible Micro-Educational Environments which the students mayexperience is delineated. Micro-Educational Environ-ments are defined in terms of the microcomputer con-figuration (e.g., hardware and software), specialeducational administrative arrangement (e.g.,resource room, special day class, mainstream), stu-dents' characteristics, and teacher attitude andtraining. Micro-Educational Environments may also bedefined by the administrative contextual variableswnich affect these environments. This description ofMicro-Educational Environments is followed by ananalysis of microcomputer effects within theseenvironments, with analyses of student-microcomputerinteractions setting the stage for studies on micro-computer effectiveness within these contexts.

RESULTS Our major research findings address how mildly handi-capped students gain access to computers and how theyspend their time while they are on the computer. Ourfindings focus both on the nature of the Micro-Educational Environment (MEE) and student use pat-terns within these environments. Overall, it isapparent that while the newness of the technology hasresulted in little consensus with regard to softwarewith a focus on math drill and practice programs,students are highly engaged by the media. Neverthe-less, the appropriateness of this instruction, andits effectiveness for cognitive and affective growthin mildly handicapped students, remains in question.

156

General findings for Year One include:

1. For LH students the main access to micros is inRS and SDC programs; access for mainstream studentsis split between the lab and classroom.

2. More "off-brand" micros in RS and SDC classes.

3. More individual use in RS/SDC programs; moregroup use in mainstream.

4. High percentage of time "on-task" across allsettings.

5. Little contact with teachers during use; contactappears to increase with the "restrictiveness" of thesetting.

6. When peers are present, LH students are morefrequently off-task with peers in RS/SDC settings.

7. Individual, small group, and whole class usereported.

8. Teachers' intentions for use varied by setting:RS-individual remediation; SDC - drill; Mainstream-recreation, reward.

9. Math software most frequently used in all set-tings (Reading almost as frequent in RS programs,followed by spelling).

10. Drill and practice type software most frequentacross all settings (>90%); "Learning to use compu-ter" type software use declines as restrictivenessincreases.

11. No commercial package used with high frequency(110 teachers reported 307 titles; 67% of titles wereunique).

12. Administrators mainly believed that motivation,not improved achievement or rate of learning, wouldbe the prime benefit of micros for LH students -across all settings.

13. Only 50% of special education programs ownedmicros; 51% of those owned were purchased usinggeneral fund (32% from federal sources).

14. Purchase decisions typically involved a districtcommittee; allocation decisions typically involved aspecial education administrator, principal andteacher.

15. Mainstream teachers more frequently had compu-ters located in classroom all day; RS teachers splitbetween class and mobile carts; SDC split betweenclass, labs, and carts.

16. Teachers perceived as having "intro" level com-puter skills by administrators; 2/3 districts repor-ted recent training; most training provided by regu-lar district staff.

FUNDING U.S. Department of Education, Office of Special Edu-SOURCE cation and Rehabilitative Services/Special EducationPrograms

RESEARCHER Melvyn I. SemmelProfessor and Program LeaderSpecial EducationGraduate School of EducationUniversity of CaliforniaSanta Barbara, California 93106805-961-3477

SPURLOCK

STUDY Special Education Work.eStation Design and Evaluation

SUMMARY

FUNDINGSOURCE

RESEARCHER

The Special Education Work.eStation project of GeorgiaTech's PhysioAutomation Laboratory is being conduc.eted in cooperation with the Georgia Computer Program.mer Project that is sponsored by Goodwill Industries.The research objective is to improve the functionaland cost effectiveness of universal teaching/trainingwork stations for severely physically disabled per.sons. Methodology involves the application of state.eof.ethe-eart computer and robotics technology, togetherwith advanced humanfactors engineering concepts, toachieve a highly facilitated integrated workstationsystem. Examples of subsystem features currentlybeing investigated are:

1. Menu.edriven touch.escreen, voice.econtrolled orpneumatically.tswitched selection of computer operations and microprocessor.econtrolled adjustment of theworkstation's mechanical features;

2. Miniature robotic device (also menu.edriven by avariety of user input signals) to supplement or eub.estitute for manual dexterity by presenting readingmaterial, drinking fluids, snacks, or facilitativedevices to the trainee; turning pages of a book ormagazine, etc.;

3. A robotic multi.esurface rotating table, posi.etioned conveniently at one end of the work.estation'smain table surface and command or signal.econtrolledby the trainee to select reading material, notepads,etc.

Through emphasis on economical and flexible designs,and the extensive use of offthe.eshelf commercialelectronic subsystems (such as voice control modulesfor personal computers), the work ..station system islikely to experience widespread use.

Georgia Tech and tLe State of Georgia, Department ofHuman Resources (Vocational Rehabilitation)

Jack M. SpurlockAssociate Vice.President for ResearchCentennial Research BuildingGeorgia TechAtlanta, Georgia 30332404-0394-e4826

STUDY

SUMMARY

FUNDINGSOURCE

RESEARCHER

a

STEELE

Development of a Camera for Use in Sensory Aids forthe Blind

The task is to determine how contemporary solid-statetechnology can be utilized in an integrated design toproduce a reading aid for the visually impaired whichis useful, portable, durable, and affordable. Thefinal device is to comprise (a) an imaging unit in ahard-tracked camera with bundled optics and illumina-tion; (b) a microprocessor to image processing andoptical character recognition on the captured data;and (c) an interface board to allow the results to besent to commercially manufactured devices for dis-play. Current research focuses on questions of tech-nical implementation, with custom equipment beingbuilt to demonstrate feasibility of alternativeapproaches, their advantages and disadvantages.

A final device of this sort should find applicationnot only in sensory aids for the visually impairedbut also in reading smediation programs for learningdisabled persons as a training tool, and in rehabi-litation programs for dyslexic adults. Results sug-gest VLSI design is well suited to meeting antici-pated demands.

VA Merit Review Approved Proposal

Richard SteeleResearch Health ScientistRehabilitation R&D CenterMail Stop 153VA Medical Center3801 Miranda AvenuePalo Alto, California 94304415-493-5000 ext. 4481

160

STUDY

SUMMARY

FUNDINGSOURCE

RESEARCHER

STEELE

Development of a Computer-Aided Visual CommunicationSystem for Aphasic Individuals

The task is to develop a useful computer-based visualcommunication system to provide severely aphasicindividuals with a means for effecting simplefunctional communication. The approach is to utilizethe most recent technical advances in microprocessordesign, computer graphics, interface devices, andaphasic research to design a system which draws onthe interact residual abilities of aphasic persons.Research questions being addressed include: (a) howshould such a system be configured; (b) what subpopu-lation of aphasics can benefit most from thisapproach; (c) how can they be most quickly identi-fied; (d) what is the most effective training regi-men? A preliminary laboratory system has beenassembled and the first subjects are now in training,with usefulness in several regards demonstrated.

A system of this sort could be useful in aphasiatherapy as well as in long-term aphasia remediation.It should help aphasiologists and neurologists tospecify more precisely the nature and extent of com-municative function that can coexist with severeaphasia. It could also be useful in training regi-mens for some severely learning disabled or otherwisedisturbed children.

VA Rehabilitation R&D Center Core Funding

Richard SteeleResearch Health ScientistRehabilitation R&D CenterMail Stop 153VA Medical Center3801 Miranda AvenuePalo Alto, California 94304415-493-5000 ext. 4481

STUDY

THORKILDSEN

An Experimental Test of a Microcomputer/VideodiscProgram to Develop the Social Skills of Mildly Handicapped Elementary Students

SUMMARY This study determined the effectiveness of a videodiscbased social skills training program for mildlyhandicapped elementary children.

The program teaches children how to use appropriatephrasing, intonation, and body language in suchsocial interactions as getting involved and beingpositive. The videodisc is used to present (1)examples of appropriate and inappropriate socialbehaviors, and (2) models to imitate in subsequentrole playing activities. A daily lesson guide forthe teacher accompanies each videodisc presentation.The program also includes a behavior managementsystem that is used during and after the videodiscand roleplaying phase.

Six elementaryschool resource rooms, each containingfive mildly handicapped students, were randomlyassigned to participate in the program (experimentalgroup) or to continue their regular resource roomprogram (control group). The students were classified as neglected, accepted, or rejected. Data onthe student's social b havior, acceptance by nonhandicapped peers, selfesteem, and treatment implementation were collected over a 4month period.

RESULTS Experimental group students scored significantlyhigher on a posttraining measure of peer acceptancethan did control group students. Also, within theexperimental group the neglected and acceptedstudents scored higher than did rejected students.The experimental and control group students did notdiffer on the postmeasure of selfesteem.

An effective use of microcomputer/videodisc technology was demonstrated and a social skills trainingprogram was validated.

FUNDING U.S. Office of Education, Office of Special EducationSOURCE and Rehabilitative Services/Special Education Pro

grams

RESEARCHER Ron ThorkildsenDirector of Technology and Administrative ServicesDevelopmental Center for Handicapped PersonsUtah State University, UMC 69Logan, Utah 84322801-750-1999

TURNER

STUDY Design and Fabrication of VOCAs at the School Dis-trict Level

SUMMARY The question guiding this inquiry was: Can thedesign and fabrication of low cost "entry level"VOCAs (and other computer aids) for low incidencespecial education populations be supported at theschool district level?

Over the past 7 years, Wayne County IntermediateSchool District (Wayne/Detroit, Michigan) has sup-ported a program of design and fabrication of lowcost VOCAs and other computer aids.

The consultant for Special Services, initially fundedunder a Title I project, is now a full time positionin the district's Data Processing Department.

Funds for the VOCAs are recycled through the resaleaccount, i.e. the devices pay for themselves via theresale charge purchasers bear, which consists ofcomponent cost and a labor cost of approximately $5-8 /hour. All other costs are borne by the interme-diate organization.

WCISD currently offers four products:

1. WOLF: nicad-powered microcomputer (65CO2) andvoice synthesizer (SC01) in Texas InstrumentsTouch and Tell case: touchpanel VOCA with 30levels, 475 words user-specified custom vocabu-lary programmed into (E)PROM. (Vocabularydevelopment system on Apple computer, usesVotrax Type and Talk synthesizer): $275.

2. ZWOLF: nicad-powered microcomputer grafted ontooriginal electronics in Tiger Electronics "K28Talking Learning Computer": used as VOCAperipheral for bank switches, Zygo scanners:$275.

3. Adaptive Input Card: similar to Schwejda Adap-tive Firmware Card, but minus several inputmodes (e.g., single switch self-scan). Set-upsoftware allows user to author ASCII and spokenvocabularies organized into menus for PowerPadtouchpanel Zygo scanners, Votrax type and talk:$150 (card and software)

4. ACES: PowerPad Touchpanel and Votrax Type and

163

RESULTS

FUNDINGSOURCE

RESEARCHER

Talk linked by menus, user programmable.Authoring and editing software; simple tutorialmode: $150 (Software)

The electronics for these devices are designed in-house, prototyped and tested and made into printedcircuit boards. In the past, enclosures and inputpanels were also custom made; currently case andtouchpanel components are obtained commercially.Fabrication is currently in-house but contractedmanufacture is being explored.

There are currently 75 WOLFS and ZWOLFs in the field.The reaction to WOLF is quite favorable and enthusi-astic. It is clear that this product meets the needsof low incidence special education populations pre-viously unexposed to VOCAs and other aids because ofhigh cost, As the price of VOCAs drops, educators aremore willing to "take a chance" with these devices.What is still needed are curricula that make moreuse of the VOCAs.

Wayne County Intermediate School District

Greg TurnerData Systems ConsultantData ProcessingWayne County Intermediate School District33500 Van Born RoadWayne, Michigan 48184313-467-1415

164

STUDY Preschool Assessment Techniques

SUMMARY

FUNDINGSOURCE

RESEARCHER

WOLFE

My intention is to use microcomputers to presentchallenges to preschoolers which would require avariety of gross motor responses (each an expectedresponse to a particular visual or aural stimulus).The response time would be measured and recorded sothat learning curves could be printed for each ses-sion, each group of sessions on the same task, andcomparisons between subjects over similar timeperiods. One initial objective is to distinguishbetween developmentally delayed and organicallydefective subjects in a systematic manner without theuse of inventories (which are costly and slow toadminister and subject to misinterpretation),

. At thesame time the experimental equipment -- dials, knobs,handles, levers, wheels, and keyboards -- could betraining gross motor skills. Diagnosis and thera-peutic efficacy could be automatically monitored andshown visually over extended time periods.

Personal Resources

Michael WolfeConsultantThe Woodlawn Center118 -13th StreetLogansport, Indiana 46947219-722-3372

165

ZUCKERMAN

STUDY Microcomputer Software for Individually Managed In-struction

SUMMARY The purpose of this project was to develop an author-ing system that would allow teachers with little orno programming experience to design courseware forcomputer-assisted instruction of handicapped stu-dents.

The research questions focus on the development ofthe product. The alpha testing of the SPE.EDAuthoring System was a formative evaluation procedureand was viewed as an integral part of the softwaredevelopment process. The overall design of the soft-ware was conceptualized, an IPO analysis and a"three-level analysis" that served as the softwaredesign reference were developed. As routines werecoded, alpha testing was used to detect logic andcoding errors. As routines were integrated, alphatests revealed any discrepancies in the conceptualdesign and the actual performance of routines. Thealpha testing indicated that the access routines andother elements of the software were consistent withthe design concept.

Questions of product impact on handicapped studentsand their teachers will be undertaken after theproduct is disseminated in final form. In the BetaTest, information about the performance of the SPE.EDAuthoring System will be obtained. The Beta Testwill be conducted in two waves. In the first wave,subjects will be "special interest, special skilledusers" who an provide commentary and feedback on thetechnical aspects of the authoring system. Subjectsin the second wave will represent the target audiencefor the authoring system. Characteristics of theseusers include rudimentary knowledge of computers, andminimal programming skills. One of the key factorsand issues in conducting beta tests is to maintain adistance from the individuals who are using the soft-ware and hardware products developed. Another issuethat arises is when should investigators contact betatest participants in order to elicit the response?

FUNDINGSOURCE U.S. Department of Education, Office of Special Edu-

cation and Rehabilitative Services/Special EducationPrograms

RESEARCHER Robert A. ZuckermanAssociate Professor401 White HallKent State UniversityKent, Ohio 44242216-672-2477

167