Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
DOCUMENT RESUME
ED 073 428 24 CS 000 377
AUTHOR Arnold, Richard C.TITLE A Comparison of the Neurological Impress Method, the
Language Experience Approach, and Classroom Teachingfor Children with Reading Disabilities. FinalReport.
INSTITUTION Purdue Research Foundation, Lafayette Ind.SPONS AGENCY National Center for Educational Research and
Development (DBEW/OE), Washington, D.C. RegionalResearch Program.
BUREAU NO ER -1 -E -023
PUB DATE Jun 72GRANT OEG-5-71-0027(509)NOTE 49p.
EURS PRICE MF-$0.65 HC-$3.29DESCRIPTORS *Language Experience Approach; *Reading Difficulty;
Reading Instruction; *Reading Research; *RemedialReading; Remedial Reading Clinics; Remedial ReadingPrograms; Tutoring
IDENTIFIeRS *Neurological Impress Method
ABSTRACTThis study investigated the effectiveness of the
neurological impress method (NIM), the language experience approach(LEA), and classroom teaching as remedial reading treatments fordisabled readers. Subjects referred to the Purdue Reading Clinic werescreened to determine whether they met criteria for the study andwere randomly assigned to the NIM, LEA, or control (classroomteaching) treatment. Tutors were also randomly assigned to either theNIM or LEA treatment. They were trained in the experimental and otherinstructional methods and were supervised in their use of remedialprocedures. Reading progress during one semester or ::;ummer sessionwas measured with oral and silent reading subtests. An analysis ofcovariance revealed that for all twelve reading measures nostatistically significant differences were found among treatmentgroups. -Growth in reading, however, was evidenced in all treatmentgroups. (Author/TO)
A,41
1Ht LT,()%
FINAL REPORT
PROJECT NO. 1- 702
GRANT NO. 0E675-71-00 7( 9_
A COMPARISON OF THE NEUROLOGICAL IMPRESS METHOD. THE
LANGUAGE EXPERIENCE APPROACH. AND CLASSROOM TEACHING
FOR CHILDREN WITH READING DISABILITIES
RICHARD D. ARNOLD
PURDUE RESEARCH FOUNDATION
PURDUE UNIVERSITY
LAFAYETTE, INDIANA 47907
JUNE 1972
U.S. DEPARTMENT OF
HEALTH. EDUCATION AND WELFARE
OFFICE OF EDUCATION
BUREAU OF RESEARCH
Final ReportProject No. 1-E-023
Grant No. OEG-S 71-0027(509)
A CCWARISON OF THE NEUROLOGICAL IliPRESS METHOD, THE
LANGUAGE EXPERIENCE APPROACH, AND CLASSROO TEACHING
FOR CHILDREN WITH READING DISABILITIES
Richard D. Arnold
Purdue Research Foundation
Purdue University
Lafayette, Indiana 47907
June 1972
The research reported herein was performed pursuant to agrant with the Office of Education, U.S. Department ofHealth, Education, and Welfare. Contractors undertakingsuch projects under Government sponsorship are encouragedto express freely their professional judgment in theconduct of the project. Points of view or opinions stateddo not, therefore, necessarily represent official Officeof Education position or policy..
U.S. DEPARTMENT OFHEALTH, EDUCATION AND WELFARE
Office of EducationBureau of Research
TABLE OF CONTENTS
CHAPTER
Page
vii
I. INTRODUCTION AND RELATED LITERATURE . 1
Introduction. . .. 1Related Literature. . . . 6 4 1Summary . . . . . . . . . . . 5Rationale for the Present Study 5
II. THE PROCEDURES AND THE EXPERIMENTALDESIGN. . . . 6 6
Sample. . . . . . . . . 6Screening Procedures. . . . . . . . . . 6Criteria for Inclusion in the Study . 6Selection and Assignment of Subjects. . 7Selection and Training of Tutors. . . . 8Instrumentation . ; . *et . 8Treatments. . . . . . . . . . . . . 11Experimental Design . . **0 0 0 h . 13Null Hypotheses . . . 13Summary . . . . . . . . 14
III. FINDINGS AND ANALYSES _ 15
Determination of Covariates . . . . . . 15Computer Program for Posttest Scores. 15Analyses of Posttest Scores 15Description of Pre Posttest Difference
Scores. . . .. . 30
Summary 30
IV. CONCLUSIONS AND RECOgHENDATIONS . . 32
Conclusions . . .. 32Discussion. . 32Limitations . . . . .. e. 33Educational Implications. . . . 33Recommendations . * . 34
-REFERENCES. . 35
APPENDIX A. . .. .. . 39
APPENDIX B . . 41
iii
LIST OF TABLES
TablePage
1. Criteria for Classification as DisabledReaders . . ..
. 7
2. Distribution of Clients Accepted for Study. 8
3. Ho.rs of Treatment Rec.ived by Each Group . 12
4. Correlation Coefficients of Pre-TestVariables . . .. . . . . . * 16
Descriptive Statistics for New DevelopmentalReading Tests Reading Diagnostic PosttestScores.
6. Results of Analysis of Covariance for ReadingDiagnostic Tests Oral Reading Posttest RawScores. . . . . . . . . . . .
17
17
Descriptive Statistics for New DevelopmentalReading Tests Vocabulary Posttest GradeScores. .
. . . . . 18
Results of Analysis of Covariance for NewDevelopmental Reading Tests Vocabulary Post-test Grade Scores
. . . . . . . .
Descriptive Statistics for New DevelopmentalReading Tests Average Reading Posttest GradeScores. . . . . . .
10. Results of Analysis of Covariance for NewDevelopmental Reading Tests Average ReadingPosttest Grade Scores
. . . . .
18
19
11. De criptive Statistics for New DevelopmentalReading Tests Literal Comprehension PosttestGrade Scores. . * M 0 0 20
12. Results of Analysis of Covariance for NewDevelopmental Reading Tests LiteralComprehension Posttest Grade Scores .
13. Descriptive Statistics for New DevelopmentalReading Tests Reading for Information Post-test Grade Scores
.
20
21
14. Results of Analysis of Covariance for NewDevelopmental Reading Tests Reading forInformation Posttest Grade Scores .. 21
iv
TablePage
15. Descriptive Statistics for New DevelopmentalReading Tests Reading for RelationshipsPosttest Grade Scores .
16. Results of Analysis of Covariance for NewDevelopmental Reading Tests Reading forRLLationships Posttest Grade Scores .
17. Descriptive Statistics for New DevelopmentalReading. Tests Reading for InterpretationPosttest Grade Scores . . . .
18. Results of Analysis of Covariance for NewDevelopmental Reading Tests Reading forInterpretation Posttest Grade Scores.
19. Descriptive Statistics for New DevelopmentalReading Tests Reading for 'Tpreciation Post-test Grade Scores
. . ..
20. Results of Analysis of Covariance for NewDevelopmental Reading Tests Reading forAppreciation Posttest Grade Scores 24
Descriptive Statistics for New DevelopmentalReading Tests Creative Comprehension PosttestGrade Scores. . . . . . 25
22
22
23
24
22. Results of Analysis of Covariance for NewDevelopmental Reading Tests CreativeComprehension Posttest Grade Scores . 25
23. Descriptive Statistics for New 'DevelopmentalReading Tests General Comprehension PosttestGrade Scores. . . . . 006 26
24. Results of Analysis of Covariance for NewDevelopmental Reading Tests General-Comprehension Posttest Grade Scores 26
25. Descriptive Statistics for New DevelopmentalReading Tests Comprehending Significant IdeasPosttest Grade Scores 27
26. Results of Analysis of Covariance for NewDevelopmental Reading Tests ComprehendingSignificant Ideas Posttest Grade Scores . 27
TablePage
27. Descriptive Statistics for New DevelopmentalReading Tests Comprehending SpecificInstructions Posttest Grade Scores . 28
28. Results of Analysis of Covariance for NewDevelopmental Reading Tests ComprehendingS .=;cific Instructions Posttest Grade Scores. 28
29. Summary of Dependent Variable Posttest WeanScores, by Treatment Groups.
0 29
30. Pre- Posttest Difference Scores, by TreatmentGroups .
. . . .. 31
vi
SUMMARY
The purpose of this study was to investigate theeffectiveness of the Neurological Impress Method (NIM),the Lank. cage Experience Approach (LEA) and classroomteaching as remedial reading treatments for disabledreaders.
The NIM is reportedly a remedial technique involv-ing multisensory learning modalities. Evidence of itseffectiveness is extremely limited. Although the LEAhas been used for many years as a remedial technique,empirical evidence is conflicting.
Subjects referred to the Purdue Reading Clinicwere screened to meet criteria for the study and wererandomly assigned to the NIM, LEA, or Control treatment.Tutors were also randomly assigned to either the NIM orLEA treatment. They were trained in the experimentaland other instructional methods and supervised in theiruse of remedial procedures. Reading progress during onePurdue semester or summer session was measured with oraland-silent reading subtests. Analysis of covariance wasused to test the statistical significance of differencesin scores attained by the three groups.
Results of analysis of covariance revealed thatfor all twelve reading measures no statistically signicant differences were found among treatment groups.Growth in reading, however, was evidenced in all treat-tent groups. It can be concluded that neither the NINAnor the LEA nor the control treatment was a more effec-tive instructional procedure than either of the others.
Lie primary implicatiJn of this studis that fordisabled readers in general, there is probably no onebest method of instruction. This implication is con-sistent with results of many studies involving typicalschool populations.
vii
CHAPTER I
INTRODUCTION AND RELATED LITERATURE
Introdt2tion
The purpose of this study was to investigate theeffectiveness of the Neurological Impress Method, whencompared to the Language Experience Approach,'ferre-medial reading instruction. A control group receivingno experimental remedial reading treatment was alsoused.
The Neurological Impress Method recently developedby Heckelman (1966) purportedly is a remedial techniqueinvolving multi-sensory learning modalities. It involvesunison oral reading by therapist and client to providean accurate model of oral reading and reinforcement ofthe client's word recognition. The index finger of bothis used to follow the print to facilitate focus and visu-al tracking and to provide large muscle emphasis for pat-terns of eye muscle movement.
The Language Experience Approach involves a client'sdictating a.reaction to an experience, the therapist'srecording the reaction in the client's own language, andthe client's reading the written form of his account ofthe experience.
An advantage claimed for the Neurological ImpressMethod is Ilat the multi-sensory approach will help thechild receive a stronger signal for learning through avisual-auditory-kinesthetie feedback system, thereby ef-fectin[- a neurological change. Advantages of the Lan-guage Experience Approach are that the language of thechild promotes interest, the words he supplies have mean-ing for him when he decodes them, the stress is on mean-ingful thought units, and word-by-word oral reading isdiscouraged.
Related Literature
Sipificance of the Problem. Until recently it hasbeen estimated fhat7TIEEled readers comprised approxi-mately ten to fifteen percent of the school population
(Harris, 1961). More current reports indicate that one.in every four students has a significant reading diffi-culty. About half of the unemployed young people todayare functionally illiterate (Grannis, 1969).
Studios (Arnold, 1969; Robinson and Hanson, 1968)have substantiated Coleman's (1940) early findings thata positive relationship exists between socio - economicstatus .nd academic success of studunts Tt is evidentthat reading disability is associated with low socio-economic status. The severity of the problem demandsthat effective remedial procedures be developed to helpalleviate the widespread and unfortunate results of lackof learning.
The Neurological Wress Method. 1.1 the early1960'sR7 CTTRWFVFITEEeve-iopeafEFNeurological Im-press Method (NIM) for children with reading problems.In an article (Heckelman, 1966) that carefully describesthe proper use of the method, it is claimed that themethod makes economical use of time and .money for mate-rials and is highly effective as a remedial technique.
Evidence of the effectiveness of the NIM is ex-tremely limited. Gains in reading were reported (Miller,1969) on a case study of one nine-year-old boy. Themethod has been reported effective with-certain cliniccases (Kaluger and Kolson, 1969). Heckelman reports onlater cases "all of the remedial students seen . . .
were able to increase their reading ability about threegrade levels or more within three months." (Heckelman,1969, p. 278).
As a result of his success in a clinical setting,Heckelman conducted a "controlled experiment" (Heckelman,1969, p. 279). He used the NIM on 24 students in grades7-10. The subjects were of average intelligence butwere ftu_ctioning three year-; or more below their expectedreading capacity. Each child received the NIM for 15 min-
,.per day, 5 days per week, for 6 weeks. With 7 1/4hours of instruction, a mean gain if 1.9 years was re-ported.
The impressive results reported by Heckelman areseriously flawed by a basic error in research design.The lack of a control treatment strongly suggests thatthe regression effect (Campbell and Stanley; 1963) waspresent and may have accounted for a good deal of the"growth" reported.
Hollingsworth (1970) developed an adaptation ofthe Heckelman /NUM to eliminate the perceived problems
2
of equiring much teacher time to help several studentsand causing voice fatigue for the teacher. In theHollingsworth experiment six children from fourth gradewere divided into matched groups, one experimental andone control. The mean scores on reading pretests wereat grade level, and intelligence scores were within theaverage range. The auditory reinforcement for the ex-perimon-al group was provided by tapes used in an E.F.I.Wirelos- System. Each chit- read 10 storms a grade be-low his measured reading grade level, 10 stories at hismeasured level, and 10 stories a grade above his meas-ured reading grade level. The teacher monitored the 8children by plugging her headset into individual receiv-ing sets. This.monitoring was reported as quite success-ful. No treatment was described for the control group.After 7 1/2 hours of impress method work by the experi-mental children, both groups were posttcsted with analternate form of the reading pretest. Analysis of vari-ance revealed no significant differences between groupson the posttest. Possible explanations were the needfor a 1:1 teacher-student ratio and the greater effec-tiveness of the NIM with remedial cases than in normalclassroom use,
The Language Experience Approach. The LnnguageExperinte ApproaCh-TEETTHES' been used for many yearsin regular classrooms. The results-of empirical researchhave been conflicting as to the efficacy of the methodfor developmental reading.. Hahn (1967) reported signifi-cantly higher scores for the LEA than for the basalreader program, on the Word ReaJing Test of the StanfordAchievement Test. Bond and Dykstra (1967) found few sig-nificant differences between the results of LEA and basalreader methods. The differences generally suported the'former program but were of little practical significanceas shown by reading achievement. In contrast, Harris andSerwer (1967) reported a slight but significant advantagefor the, basal reader progra 1 as 1 of 4 met 'Asp includingLEA, UE0'-'. in a compc.rcltiv: study. Neither r'Jid andPeltramo (1966) nor Robinson (1968) support one methodover another.
Spieth (1967, 1968, 1969) reported. on three experi-men ,1 summer remedial reading programs, RISE I, RISE II,and RISE III, for first and second graders. The reportsdid not indicate whether any children participated inmore than one RISE program. The programs compared meth-odologies based on LEA using Words in Color; Scott,Foresman basal readers; SulliVEHTraFraEFFE readers; andSPA Readi Laboratories, for effectiveness in increasing, Laboratories,
reading ability, improving behavior in the classroom, in-creasing attention span, and improving eye-hand perceptu-al skills. Classes of fifteen children met for fourhours a day.
In RISE I differences in achievement .were not greatamong groups taught by different methodologies. LEA pro-vided f-eater numerical galls in measured reading abil-ity than any other method. Significant advantages (p.05)for LEA ;appeared in the development of good behavior inthe olaOroom, improvement in perceptual skill, and in-crease of attention span (Spieth, 1967).
The children's personality types were studied andcategorized as problem behavior, lack of perceptualskills, shyness and withdrawal, or combinations of prob-lems. Among the withdrawn children, second graders im-proved in reading more than first graders (p.05). Firstgraders made more progress than second graders in class-room behavior, perceptual skills, and attention span.LEA was the most effective methodology for withdrawnchildren, particularly for increasing attention span(p.05) (Spieth, 1967).
RISE II (Spieth, 1968) was to test the .usefulnessof grouping children by personality types in order toprovide every child with the teaching methodology shownby RISE 1 to be most effective. LEA was the methodologyused with withdrawn children and with those having sev-eral problems. For the latter group of children, sig-nificant improvement over subjects in RISE I was reportedfor digit span and Bender-Gestalt Test results. Secondgraders were signifTEWITTFifEJF MT their counterparts-of RISE I when_average reading 'scores were compared. Gen-erally, the expectation of better learning as a result ofappropriate instructional method for each child was notfulfilled. A third program was conducted to try to cor-rect tl i lack of progress clidenced.
For RISE III, children of different personalitytypes were deliberately mixed in the same classroom.Groupings for instruction were formed within the class-room on the basis of personality type, and children ineach group were taught by the methodology found most ap-propriate in the earlier program. This arrangement.seemed to provide the most satisfactory learning condi-tions. "All comparisons between RISE II and RISE IIIwere significant-at least at the .05 probability leveland in the predicted direction" (Spieth, 1969, p. 102).
4
NeasureL, reading growth of children using LEA was greaterthan the measured growth of children using other methods.
Grace Fernald -(1943) reports'using a technique verysimilar in many ways to LEA. In her study remedial cases.reportedly made great strides in reading Rrowth. It isgenerally accepted that the LEA in clinical settings-hasbeen an effective-method for treating reading disabilitycases (Harris, 1970).
Summary
Very few studies of-the use-of the Neurological Im-press Method (NIN) have been reported in the literature.Two reports were authored by the developer of the-method,and a- basic flaw in the design is considered a limita-Lion of the research validating the method. One re-port of a study of Children from typical classrooms re-vealed possible limitations from lack of one-to-oneteacher - student use of the method. No differences betweentreatment-and control groups were observed.
Reports of the Language Experience Approach (LEA)studies showed tonflicting'results from its use as an in-structional technique in the classroom_ Its use as a re-medial technique has been seldom reported but widelyaccepted.
Rationale for the Present Study: -
For this study it was felt that'the use of controlsubjects in community schools and two experimental meth-ods in the Reading Clinic would provide multiple compari-sons of effectiveness of methods, and would eliminatethe wea:rnesses of previous 3tudies.
On the basin of previous studies, differences ingrowth among NM, LEA, and Control treatment groups inthis study were hypothesized for oral reading and vari-ous reading subtests.
5
CHAPTER I
THE PROCEDURES AND THE EXPERIMENTAL DESIGN
Sample
The population for this study was drawn from asmall city and adjacent rural communities within commu -ing distance of Purdue University, The socio-economiclevel is predominantly middle class, but all classes-arefound in the area.
All subjects were referred to the Purdue UniversityReading Clinic. The subjects considered for inclusion inthis study were regular students in community schools.Their range of grade placement was from first gradethrough high school. The population included many moremales than females.
Screening Procedures
All clients referred to the Reading Clinic fromJuly 1970 through September, 1971, were screened for in-clusion in the sample for this study. After their par-ents had completed appropriate application blanks andhad signed permission forms (Appendix A) for their childto participate in research, subjects visited the ReadingClinic, and a clinic supervisor administered the screen-ing instruments.
Criteria for Inclusion in the Study
Three criteria were set-for acceptance into thestudy:
1. That the subjects. attain an intelligence quotient o80 or higher on either the Peabody Picture VocabularyTest (PPVT) (Dunn, 1959) or Raven's ProgressiveMatrices (Raven's) (Raven, 1965).
That the subjects be considered serious reading dis-ability cases, as indicated by the difference betweenthe obtained grade score on the reading subtest of theWide Range Achievement Test (WRAF) (Jastalc, 1965) and
6
the .child's actual grade placement. The necessarydifferences for various grade placements, recom-mended by Kaluger and Kolson (1969, p. 48), can beseen in Table 1.
That the subjects have no known primary disabilitysuch as emotional disturbance or neurological defi-cit' which would seriously impair learning.
Table 1
Criteria for Classification asDisabled Readers
Grades Behind in Reading
1,2
3,4
5,6
Junior High
Senior High
3 -6 months
6-8 months
9 months - 1 year
1 year - 1 1/2 years
1 1/2 years 2 years
Selection and Assignment of Subjects
A total of 63 children were screened for this study.Of these 63 candidates, 23 children were not included inthe study because-they did lot meet the criteria for ac-ceptance. Those 40 who met the established .criteriawere then randomly assigned to one of three treatments.
The distribution of clients accepted for the studyis shown in Table .2. Due to various reasons, incompletedata were obtained for eight subjects. Thus a total of32 subjects furnished the data repotted in this study.Ten subjects completed the NIM treatment, thirteen sub-jects completed the LEA. treatment, and nine subjects com-pleted the Control treatment.
7
Table 2
Distribution of Clients Accepted for Study
NII LEA Control
Total Admitted and Assigned
15 15
Total T
10 13
aced
10
9
Selection and. Training of Tutors
Every subject included in this study received re-medial teaching from one assigned tutor.
Tutors used in this study were students enrolledin Education 536, Foundations. of Remedial Reading, thefirst course in a three-course sequence to prepare re-medial reading teachers. As _a part of the course re-quirements, students participate in a three hour labora-tory experience in the Redding Clinic, tutoring a childthree days a.week, Tp eliminate any possible bias of tu-tors, theywere randomly assigned to either the NIM or theLEA condition. Since Control subjects were not seen atthe Reading Clinic, no tutors were involved in the Con-trol condition.
All tutors were trained in both experimental pro-cedures and instructed to use only that procedure towhich they were assigned. They were further instructednot to use any procedure which might conflict with theirassigned method. Tutors were supervised and told to askfor help if a particular activity was doubtful to them.
Instrumentation
Every subject in this study was administered fivestandardized measures. Three measures were administered
by a clinic supervisor for screening. purposes. Two cri-terion measures of pre-post experimental gain were admin-istered, one in oral reading by a clinic supervisor, andone for silent reading by the assigned tutor.
Screening Measures. The Wide Range AchievementTest On CEstn, JaStak, and Bijou, 1965) is an indi-vidual test of reading (word recognition and pronuncia-tion), 4T tten spellinc al,d arithmetic co_iputation.Its purpose, according to tie manual, is to provide anassessment of achievement as an adjunct to intelligenceand behavior adjustment tests. The standardization sam-ple was widely distributed over the United States. Valid-ity'of WRAT scores is shown by correlations with bothschool grades (.78 to .88) and results of individual in-telligence tests (.66 to .84). Reliability coeffiCientsare given for each subtest and age group (not less than.940 for any age group) . This instrument has been favor-ably reviewed (Buros l94, 1965).
The Peabody Picture Vocabulary Test (PPVT) (Dunn,1959) is an individual test of intelligence. Its gen-eral purpose is to estimate verbal intelligence by as7sessing hearing vocabulary, The PPVT may be used withpersons from age 2.5 .to age 18.
The final standardization group included 4,012 whitechildren and youth residing in and around Nashville,Tennessee. TL .J sibjocts word c4orl to represent normaldistributions of intelligence as determined by scores onthe Kuhlmann-Finch Intelligence Test. Mean Test Age wasused to determine age norms for raw scores obtained onthe test.
Statistical validity is reported in the manual.PPVT scores increased with the age of the subject, cor-relate with Wechsler more than- with Binet I.Q.'s, andare fa=rly related. to scho. 1 achievement. Reliabilityas reported in the manual 7::_lngos from 0.67 to 0.84. Re-.viewers (Buros, 1965) have commented favorably abOut thisinstrument.
Both forms of Raven's Progressive Matrices (Rd en's)(Raven, 1960, 1965) provide estimates of nonverbal intel-ligence. Their purpose is to supplement vocabulary testswith perceptual tasks for subjects of all ages. The in--struments were standardized on British populations. De-scriptions of validity and reliability are provided inthe manuals. Reviews (Buros, 1949, 1953, 1959, 1965)stress the usefulness of this instrument for subjectswith communication difficulties.
9
Criterion 1,1easuro of Ural Readin;, The pretest-posttestT,casure er-dral-Fealang accuracy was the oralreading subtest of the Cates-ikKillop Reading Diagnos-tic Tests (RD?) (Gates, McKtllop, 1952). The same formwas used for pretest and oosttest. This instrument isa revision of t:le Gates Reading Dianestic Tests (Gates,1926) which were sta,1J-rdized on a group of third gradersof wid(ly varied hoe and r-iltural backgronnds. ThechildrLn ::e ,1 c,_t rA nr en'. New Yor City Themanual lacks specifi data on reliability and validity.The test is widely used, mil 1-.!.1 oral reading subtestappears to have face vali=".ty. This instrur:lont has beenfavorably reviewed in tie i;iental 1.4easurements Yearbook(Burros, 1949, 1953, 1959, 1t:Y-7.5).
Criterion Measures of Silent Rearing The New De-velcpmental Reding To,Hs '(.2-onct7=-LITEWTlieyt, 1968)(NDKT) were used to assess nre-post experimental gain insilent reading and comnrehensicn. This instrument hasbeen favorably review?d 1972).
The Priv=y level te, : standardized on morethan 5000 children in stratifi d rando;ized samples fromtwo large midwestern co=uni Validity was estab-lished by verification of g de placement using otherreading tests. Reliability was shown by correlation co-efficients between .89 and calculated from between-form score,
The Primry level tests were used with 21 subjectsLa this study. buhtest scores thus obtained were the de-pendent variables Vocabulary,. Comorehending Significantideas, Comprehending Specific instructions, Literal Com-prehension and Ave-2a ge ReadLar
The Into=dia level tests e standardized on15,000 pupils in lc.arnig contc)rs of wide geographicaldistriAltion ever the Ctates. validity wasshown by poit-bisciul.curvelations clustcing mostlybetween AO and .54. Reliability coefficients based oninternal consistency -ar.a,eit ,:cm .83 to .94.
The intermediate level tests vero.used with 11 sub-jects. in this stu6y. Subtest scores thus obtained werethe dependent variables Vocabulary, Reading for Informa-tion, Reading for Relations,. Literal Comprehension,Reading for Interpr,Jtation, Reai7 for Appreciation,Creative Com?rehension, rehonsion, and Aver-age Reading.
10
Treatments
Clinical Procedures. Tutors and subjects workingin theIZIUTrT CriTleTJgid their assigned experimentalmethod and ether non-conflicting methods and materialsto-remediate the subjects' reading problems.
The Neurological Impress Method (NM ) involved thetutor''; reading orally with the child as bo h pointedwith index fingers to the words being,-read. The tutorwas to read slightly faster than the.:_subject's usualoral reading rate. According to Heckelman, the tutorwould provide a model of fluent oral reading and re-inforce the subject's word recognition. The fingerswould help focus visual attention on the word being readand provide large muscle reinforcement for eye movements.No attention was given to comprehension unlesS the sub-ject asked questions, as prescribed by Heckelman (1966,1969)
The Language Experience Approach (LEA) involved thetutor's writing the subject's dictated accounts of experi-ences. These stories served as text from which theHsub-ject learned to read. The subject would learn to recovnize his own words in other settings. The child would bereading about a topic of interest to him. The readingvocabulary would be meaningful thought-unit.phrasingratherthan mere word-calling, as emphasized by Lee and Allen(1963) .
Fifteen minutes of each tutoring session were de-voted to the experimental treatment. The remaining timewas devoted to other reading tasks which were considerednon-interfering with the experimental treatment. Non-interfering tasks included such activities as buildingsight vocabulary and developing word analysis and compre-hension skilis
Aemediation was conducted at the instructional andindependent reading levels as determined by InformalReading inventories (Johnson and Kress, 1965) adminis-tered to-each subject by his assigned tutor. InformalReading Inventories were prepared from basal reading mate-rials in the Purdue Reading Clinic. Each subject readwords in-isolation and paragraphs. Accuracy-of oralreading and comprehension were evaluated to-determinegrade levels at which the subject could read independent-ly and with tutor assistance. Criteria for evaluatingperformance were those of Johnson and Kress (1965). Thefrustration reading level was carefully avoided by eachtutor.
11
Each tutor was supervised by the writer and bysupervisors working in the Reading Clinic. Tutors re-ceived individual consultations regarding their clientsduring the experimental period. Materials for instruc-tion were high interest, low vocabulary trade books,tutor-made exercises, and other materials available-inthe Reading Clinic.
Control Treatment Procedure. After the initialscreening;-T6ntrol-Treatment .S.uTects received no re-mediation in the Reading Clinic. However, these sub-jects did receive-training as part of regular classroominstruction. This treatment group was included primari-ly to determine the possible effect of regression tothe mean frequently involved in the study of below aver-age performers (Campbell and Stanley, 1964).
Treatment Period. Subjects were included in thestudy during fall7semesters, one spring semester,.and one summer session. Each subject was included inthe study for only one term.
Each subject in the NTH and LEA groups was sched-uled to attend the Purdue Reading Clinic three.,timesper week for the semester, to receive approximately 35tutoring hours in fall and spring terms and 23 hours inthe summer session. Tutoring sessions were SO minutesin length in the fall and spring, 60 Minutes in summer.
The mean number of hours of remediation for eachgroup is listed in Table 3. Most of the difference intreatment time was due to absenCe of the subject, an un-controlled factor in this study.
Table 3
lours of Treatment Received-by Each Group
Treatment NIM LEA Control(Estimates)
Total Instruction
Range 17-29 1/6 13-25 0-90
Mean 24 30 26
Experimental Treatment
Range 1/2 -8 3/4 3 1/4-10 1/2 Unknown but
Mean 7 6 1/2 robably
12
Attendance records of subjects in the NIM and LEAtreatment groups indicated the number of lessons the subjects received. Since the Control subjects were not seenin the Reading Clinic except for screening purposes, nodata were available on the exact amount of remedial helpavailable to those subjects. In order to determine thisas closely as possible, questionnaires (Appendix B) weresent to parents and teachers of the Control subjects toascertL_n the amount of spc.;ial reading he4 given duringthe experimental period. questionnaires were sentafter each treatment period and the returns were 100%.
Experimental Design
Analysis. The basic design involved a comparisonof the Control and the two experimental groups on twelvescores. Direct scores were obtained for eight subtests,and four more by combining two or more subtests. Differ-ence scores from pre- to posttest were calculated foreach of the dependent variables.
Descriptions of the range, mean, and standard de-viation of the final scores were obtained using ProgramDISTAT from the EDSTAT series (Veldman, 1969).
Determination of covariates (Elashoff, 1971) wasfrom correlation matrices obtained by using BMD 3D ofthe Biomedical Program Series from the University ofCalifornia School of Iiedicine.
The covariates determined from the correlationMatrices were Raven's IQ equivalent, WRAT spelling rawscore, and chronological age.
The principal analytical form was analysis of co-variance (Weiner, 1971) on the posttest scores. Themethod used was "BMDO 4V - Analysis of CovtrianceMultiple Oovariatos - Version of April 1, 1966," (Dixon1970) .
Null Hypotheses
The principal hypotheses of the study were as fol-lows:
Among the NIM, LEA, and Control treatment groups, whenscores are adjusted for the effects of the covariates,there will be no differences in scores of:
13
a) Oral Reading,b) Vocabulary,c) Average Reading,d) Literal Comprehension,e) Reading for Information,f) Reading for Relationships,g) Reading for Interpretation,h) Reading for Appreciation,
Creative Comprehensionj) General Comprehension,'k) Comprehending Significant Ideas, or1) Comprehending Specific Instructions.
na
Th population for this study were school childrenliving within cc_muting distance: of Purdue University andreferred to the Pv)r.due-Reading Clinic for help with read-ing problems. The subjects were screened to meet cri-teria established ftJ. the study and were randomly assignedto the Neurological Impress Method, the Language Experi-ence Approach, or the Control tl:eatment.
The tutors enrolled in the first of three graduatecourses in remedial reading, were also randomly assignedto either the NIM or the LEA treatment. They weretrained in the experimental and other instructional meth-ods and supervised in their use of remedial procedures.
Each subject was included in the study for one Pur-due semester or summer session. Reading progress of sub-jects was measured with oral and silent reading subtexts:Analysis of covariance-was used to test the statisticalsignificance or differences in scores attained by thethree treatment sroups.
14
CHAPTER III
FINDINGS AND ANALYSES
Results of this study are presented as statisticalanalyses- and descriptive statistics.
Statistical analyses are based on.both. pretestscores and posttest scores. Pretest scores were used todetermine the covariates, and posttest scores were ana-lyzed by Analysis of CovarianCe.
The first descriptive statistics provided are basedon posttest scores from the criterion measures, Gates7McKillop Reading Diagnostic Tests -(RDT) and the New De-velopmental Reading Tests (NDRT) by Bond, Balow, and Hoyt.Descriptive statistics based on pre-posttest differencescores are also included.
Determination of Covariates
Determination of covariates was based on pretestscores with correlations of .3 or higher (Elashoff, 1971).The correlation matrix presented in Table 4 was obtainedusing BMD 3D of the Biomedical Program Series from theUniversity of California School of Medicine (Dixon, 1970).The covariates so determined were Raven's IQ equivalentscore, WRAT Spelling raw score, and chronological age.
Comps ter Programs for Posttest Scores
The descriptive statistics presented. in this Sec-tion were obtained using Program DISTAT from the EDSTATseries (Veldman, 1969). Analyses of Covariance (Weiner,1971) were computed by "BMDO 4V Analysis of Covariance-Multiple Covariates - Version of April 1, 1966" Dixon,1970).
Anal- -ses of Posttest Scores
In Table S the ranges, means, and standard devia-tions of RDT Oral Reading raw scores are presented bytreatment groups. The highest mean score (11.3) was
15
TA
BL
E 4
Correlation Coefficients of: Pre.-Test
Variables
.
ITSItiftlt.
Pelted),
Raven
eniAt
VRAT
'MAT'
Carl
Path,
Age,
Pre
Vaceb.
Pre
Avg.
14.
Pre
Literal,
Camp.
PTO
Rdt. far
Inform,
t146
General
144, far
.1112. 167 444. for
Creative
Campre-
Itelmlian- 1nterpre- Afpta-
Cospreatm- benelvm
ships
tatiam
tion
eion Pre
?Ts
14g., far
Signifl-
cant
.
ideas Pi.
Sfeti
Lair .4,
tit
Pr.
sea
Treatment.
74107 141
Rime's 141.
494714"61142
41141 4411161
mir mail
At,
Gaiter .Pre
Paten. PTS,
441.-14... 974
Lit.
Camp.
044
for
::lettMetips
44g. for
larterfretatiap
161g. for
Appreciatieu
Creative
Comprehenaloe
General
Cum9444694146
Aig. for 11401f.1-
cant lies.
*minx'
lastractleas
...Ain.
7.1144
.1324-
.4054
Am
.4711
.2457
3434
.4544.
-.0662
.2124
'-..3152
.1674.
-.1441
.7769
.3095
.0124
.2561
.3374
.4543
.1117
.1422-.0744-.4414
.0145
.1421 -.1341
.7412 -.1932.
.3047
.8513.-.6236
.1734
-,12.14
.4143,
.1411
-.4329
-.0201
.3314
.19i4
.4701
-.4107
-.2117
.7128
41AI
-.024.5
....2344,
-.1555.
-.1211
-.3097'
-.4466
-.4111.
.4314
.2041,
.2754
_0945
.1654.
2414
.2351'
.-.0165
-.3653
-.2571
.6509
.2447
.:::::,
.4900
.4444
:::::
-.0401
,..7154
.1054,
.5447
:::::
.4404
.0795.
.0760
.4044
.75246
-....417:1:
.5144
.9443.
.4440
.....0411.0s2:5;:b,
.4517
-.2774
-.2112
-.1445
.1993
-.2444
.9436.
.4274
.7944
.1107
..,..:::
.4941.
-.540.6
-.,11171:
-.1448
-.1104
.7ill'
.9343
.4441
.4233
.4112'
.9324
:::::
.4772.
-..7..:79,15:4071411.31.
-43JOS
-.1725.
.14414
-.3179
.1541.
:ma*
.1147
.9774
.91171.
.04441
.4871.
.1213
.5576
-.4614.
..5849,
.3735
-.3105'
.074
4
.9152
.1141.
.9154,
.9414
.0444
.4114
4645
.4044
.1449
.1293
.1201.
.1145.
.1641
.35611
.2645
.4441
.1107
-.0213.
.
.1494
.0404.
.4044
.9000
.4044
.oaot
.4010
.20m7
.2614
.1327
.0617
.596
15
Admat,
.1115
: :4214
' .1147'
.2454
.7.94591
.4440
:0400
.61006
.4000'
.1414
.0000
.16413
attained by the Control treatment group, the socond high-est mean-score (10.4) was attained by the LEA treatmentgroup, and the lowest mean score was attained by the NIMtreatment group.
Table 5
Descriptive Statistics for New Developmental ReadinTests Reading Diagnostic Posttest Scores
Treatment NIN LEA Control
Range
Mean
St. Dev.
1.3-20.8
10.4
6.8
3.8-20.8
11.3
7.2
9
The results of Analysis of Covariance of RUT OralReading raw scores are presented in Table 6. The differ-ences in means among the treatment groups were statistical-ly nonsignificant.
Table 6
Results of Analysis of Covariance for ReadingDiaz rAic-Tests Oral Reading.
Posttest Raw Scores
Source
Treatme_
error
Degrees ofFreedom
MeanSquare.
2 518.9300 .132 N.S.
26 3922.1076
In Table 7 the ranges, means, and, standard devia-tions of the NDRT Vocabulary grade scores are presentedby treatment groups.. The higher mean score (4.5) was at-.tained by the Control treatment group, and the lower mean
17
score (4.1) was attained by both the LEA and the NIMtreatment groups.
Table 7
Descriptive Statistics-for New DevelopmentalReading Tests Vocabulary Posttest
Grade Scores
Treatment NIM LEA Control
Range
Mean
St. Dev.
1.3-8.4
4.5
2.1
9
The results of Analysis of Covariance of the NDRTVocabulary grade scores are presented in Table 8, Thedifferences in means among the treatment groups werestatistically nonsignificant.
Table 8
Results of Analysis-of Covarlance for NewDevelopmental Reading Tests Vocabulary
Posttest Grade Scores
SourceDegrees of MeanFreedom Square
Treatment
Error
2
26
268.2232 1.090 N.S.
245.9975
In Table 9 the ranges, means, and standard deviationsOf NDRT Average Reading-grade scores are presented bytreatment groups. The highest mean score (4.8) was at-tained by the Control treatment group, the second highestmean score (4.1) was attained by the LEA treatment-group,and the lowest mean score (4.0) was attained by the NIMtreatment group.
18
Table 9
Descriptive Statistics for New Developmental ReadingTests Average Reading Posttest Grade Scores
Treatment NIM LEA Control
Range 1.9-9.1 1.5-8.0 1.9-9.2Mean 4.0 4.1 4.8
St. Dcv. 2.2 1.6 2.2
N 10 13 9
The results of Analysis of Covariance of NDRT Aver-age .Reading grade scores are presented in Table 10. Thedifferences in means among the treatment groups were sta-tistically nonsignificant.
Table 10
Results of Analysis of Covariance for NewDeVelopmental Reading Tests Average
Reading Posttest Grade Scores
SourceDegrees ofFreedom
MeanSquare
Treatment
Error
2
26
333.5846 1.445 N.S.
230.8489
In Table 11 the ranges, means, and .standard devia-tions of NDRT Literal Comprehension grade scores are shownby treatment groups. The highest mean score (4.8) wasattained by the Control treatment group, the second high-est mean score (4.2) was attained by the LEA treatmentgroup, and the lowest mean score (3.6) was attained bythe NIM treatment group.
19
Table 11
Descriptive. Sta .stics for New Developmental ReadingTests Literal Comprehension PostteSt Grade Scores
Treatment NIM LEA Control
Range
le an 3.6
St. Dcv. 1.5
N 10
1.9-8.3
4.2
1.7
13
2.0-9.9
4.8
2.2
9
The results of Analysis of Covarian'ce of the NDRTLitaral.Comprehension grade scores are shown. in Table 12.The means among the treatment groups wereStatistically nonsignificant.
Table 12
Results of Analysis of Covariance for NewDevelopmental Reading Tests LiteralComprehension Posttest Grade Scores
SciurceDegrees ofFreedom
MeanSquare p
Treatment
Error
2 264.1172 1.233 N.S.
26 214.1670
In Table 13 the ranges, means, md standard devil-tions of NDRT Reading ,for Information gra4_scores arepresented by treatment grout's. The highest mean score(6.1) was attained by the NIM treatment group, the sec-ond highest mean score (5.4) was attained by the Controltreatment group, and the lowest mean score (5.2) was at-tained by the LEA treatment group.
20
Table 13
Descriptive Statistics for New Developmental ReadingTests Reading for Information Posttest Grade Scores
Treatment NIM LEA Control
Range 5.4-6.8 3.6-7.6 0.1-8.6
Mean 6,1 5.2 5.4
St. Dev. 0.7 1.6 3.8
N 3 5 3
The results of Analysis of Covariance of the NDRTReading for Information grade scores are presented inTable 14. The differences in means among the treatmentgroups were statistically nonsignificant.
Table 14
Results of Analysis of Covariance for NewDevelopmental Reading Tests Reading for
Information Posttest Grade Scores
SourceDegrees ofFreedom
MeanSquare
Treatment
Error
2
26
406.6259 2.176 N.S.
186.8400
In Table 15 the ranges, means, and standard devia-tions_of NDRT Reading for Relationships grade scores arelisted by treatment groups. The highest mean score (6.5)was attained by the Control treatment group, the secondhighest mean score (6.3) was attained by the NIM treat-ment group and the lowest mean Score (5.8) was attainedby the LEA treatment group.
21
Table 15-
Descriptive Statistics for New Developmental ReadingTests Reading for Relationships Posttest Grade Scores
Treatment LEA Control
Range 6.1-6.8 3.78.6 4.7-9.6
Mean 6.3 5.8 6.5
St. Dev. 0.4 2.0 2.7
N 3 5 3
The results of Analysis of Covariance of the NDRTReading for Relationships grade scores are listed inTable 16. The differences in means among the treatmentgroups were statistically nonsignificant.
Table 16
Results of Analysis of Covariance for NewDevelopmental Reading Tests Reading for
Relationships Posttest Grade Scores
SourceDegrees ofFreedom
MeanSquare p
Treatment
Error
2
5
274.7387 .862 N S.
318.8648
In Table 17 the ranges, means, and standard deviations of the NDRT Reading for Interpretation grade scoresare presented by treatment groups. The highest meanscore (7.4) was attained by the Control treatment group,the second highest mean score (6.9) was attained by theNIM treatment group, and the lowest mean score (5.5) wasattained by the LEA treatment group.
22
Table 17
Descriptive Statistics for New Developmental ReadingTests Reading for Interpretation Posttest
Grade Scares
Treatment NIM LEA Control
Range
Mean
St. Dev.
N
4.5-8.9
6.9
2.3
3
3.7-8.7
5.5
2.0
5
6.1-9.7
7.4
2.0
3
The results of Analysis of Covariance of the NDRTReading for Interpretation grade scores are presented inTable 18. The differences in means among the treatmentgroups were statistically nonsignificant.
Table 18
Results of Analysis of Covariance for NewDevelopmental Reading Tests Reading forInterpretation Posttest Grade Scores
SourceDegrees ofFreedom
MeanSquare
Treatment
Error
2
5
F p
506.4097 1.665 N.S.
304.1003
In Table 19 the ranges, means, and standard devia-tions of the NDRT Reading for Appreciation grade scoresare shown by treatment groups. The highest mean score(8.0) was attained by the Control treatment group, thesecond highest mean score (7.9) was attained by the NIMtreatment group, and the lowest mean score (5.3) was at-tained by the LEA treatment group.
23
Table 19
Descriptive Statistics for New Developmental ReadingTests Reading for Appreciation Posttest Grade Scores
Treatment NIM LEA Control
Range 7.1-9.5 2.7 -9.b 5.5-10.6Mean 7.9 5.3 8.0
St. Dev. 1.4 2.6 2.5
N 3 5 3
The results of Analysis of Covariance of the NDRTReading for Appreciation grade scores are shown inTable 20. The differences in means among the treatmentgroups were statistically nonsignificant.
Table 20
Results of Analysis of Covariance for NewDevelopmental Reading Tests Reading forAppreciation Posttest Grade Scores
SourceDegrees ofFreedom
MeanSquare
Treatment
Error
F p
1359.4242 3.195. N.S.
425.5118
In Table 21 the ranges, means, and standard devia-tions of the NDRT Creative Comprehension grade scores arepresented by treatment groups. The highest mean score(7.7) was attained by the Control treatment group, thesecond highest mean score (7.3) was attained by the NIMtreatment group, and the-lowest mean score (5.0) was at-tained by the LEA treatment group.
24
Table 21
Descriptive Statistics for New Developmental ReadingTests Creative Comprehension Posttest Grade Scores
Treatment N LEA Control
Range 5 8-9.0 3.2-8.9 -10.4
Mean 7.3 5.0 7.7
St. Dev. 1.6 2.3 2.3
N 3 5 3
'The-results of Analysis of Covariance of the NDRTCreative Comprehension grade scores are presented inTable 22. The differences in means among the treatmentgroups were statistically nonsignificant.
Table 22
Results of Analysis of Covariance for NewDevelopmental Reading Tests CreativeComprehension Posttest Grade Scores
SourceDegrees ofFreedom
MeanSquare F p
Treatment
Error
2 1020.7729 2.873 N.S.
355.2651
In Table 23 the ranges, means, and standard devia-tions of the NDRT General Comprehension grade scores areshown by treatment groups. The highest mean score (7.5)was attained by the Control treatment group, the secondhighest mean score (6.9) .was attained by the NIM treat-ment group, and the lowest mean score (5.0) was attainedby the LEA treatment group.
25
Table 23
Descriptive Statistics for New Developmental ReadingTests General Comprehension Posttest Grade Scores
Treatment NIM LEA Control
Range 6.1 -8.1 3.2-8.9 5.5 -10.2
Mean 6.9 5.0 7.5
St. Dev. 1.v 2.3 2.4
N 3 5
The results of Analysis of Covariance of the NDRTGeneral Comprehension grade scores are shown in:Table 24.The differences in means among the treatment groups werestatistically nonsignificant.
Table 24
Results of Analysis of Covariance for NewDeveloPMental Reading Tests GeneralComprehension Posttest.. Grade Scores
SourceDegrees ofFreedom
MeanSquare
Treatment
Error
720.0498 2.184 N.S.
329 6595
In Table 25 the ranges, means, and standard devia-tions of the NDRT ComprehendingSignificant,Jdeas-Fgradescores are listed by treatment sroups. The:highest meanscore (3.7) was attained by,the Control treatment.group,the second highest mean score (3..5),was attained ;by. theLEA treatment-group, and t.helowest mean score 3,1) wasattained by the NIM treatment group.
26
Table 25
Descriptive Statistics for New Developmental ReadingTests Comprehending Significant Ideas Posttest
Grade Scores
Treatme NIM LEA Control
Range 2 5 2 1.6-5.2 1.9 -5.2
Mean 3.1 3.5 3.7
St. Dcv. 1.1 1.3 1.2
N 7 8 6
The results of Analysis of Covariance of the NDRTComprehending Significant Ideas grade scores are listedin Table 26. The. differences in.meanS among the treat-ment groups were'Statistically nonSignifican
Table 26
Results of Analysis of Covariance for NewDevelopmental Reading Tests ComprehendingSignificant Ideas PoSttest -Grade. Scores
SourceDegrees ofFreedom
MeanSquare F
Treatment
Error.
2
15
.3255 .054 N.S.
154 8494
In Table 27 the ranges, moans, and standard devia-tions_of the NDRT Comprehending Specific Instructionsgrade scores are presented by treatment groups. Thehighest mean score,(3,8) was attained by-the Controltreatment group, the second highest 'mean 'score 13.6)was.. attained by the LEA treatment group, and the, lowestmean' score 3.1), was attained by the NIM treatment group.
27
Table 27
Descriptive Statistics for New Developmental, ReadingTests Comprehending Specific Instructions
Posttest Grade Scores
Treatmc at NIM LEA Control
Range 2.5-5.2 2.1-5.1 2.3-5.1
Mean 3.1 3.6 3.8
St. Dev. 1.1 1.1 1.1
N 7 8
The results of Analysis of Covariance of the NDRTComprehending Specific Instructions grade scores are prrented in Table 28. The differences in means among thetreatment groups were statistically nonsignificant.
Table 28
Results of Analysis of Covariance for NewDevelopmental Reading Tests Comprehending
Specific Instructions PosttestGrade Scores
SourceDegrees ofFreedom
MeanSquare
Treatment
Error
2
15
30.1140 .274 N.S.
119.9907
In Table 29 the mean posttest scores for all de-pendent variables are presented by treatment groups.The NIM treatment group attained the highest mean scoreon one subtest, NDRT Reading for Information, and theControl treatment group attained the highest mean scoresfor all other subtests. None of the differences amonggroups were statistically significant.
28
Table 29
Summary of Dependent Variable PosttestMean Scores, by Treatment Groups
VariableNIM
TreatmentLEA
TreatmentControlTreatment Value
RDTOral Reading 9.1 10.4 11.3 N.S.NDRTVocabulary 4.1 4.1 4.5 N.S.NDRTAverage Reading 4.0 4.1 4.8 N.S.NDRTLiteral Compre-hension 3.6 4.2 4.8 N.S.NDRTReading forInformation 6.1 5.2 5.4 N.S.NDRTReading forRelationships 6.3 5 8 6.5 N.S.NDRTReading forInterpretation 6.9 5.5 7.4 N.S.NDRTReading forAppreciation 7.9 5.3 8.0 N.S.NDRTCreative Compre-hension 7.3 5.0 7.7 N.S.NDRTGeneral Compre-hension 6.9 5.0 7.5 N.S.NDRTComprehendingSignificantIdeas 3.1 3.5 N.S.NDRTComprehendingSpecificInstructions 3.1 3.6 3.8 N.S.
DescriEtion of Pre-Posttest Difference Scores
The descriptive statistics for difference scoreswere obtained using Program DISTAT from the EDSTATseries (Veldman, 1569).
These scores were not subjected to analysis be-cause of statistical difficulties associated with dif-ference scores ;;egression to the mean frequently oc-curring among subjects originally scoring at the ex-tremes of an instrument's range is one such effect. Theceiling and floor effects of tests designed for particu-lar ability levels also make difference scores less use-ful-for comparing groups within a study than are post-test scores adjusted for the effects of covariates. Meandifference scores are included here for comparison withother studies.
In Table 30 the mean difference scores for all de-pendent variables are presented by treatment groups.The NIM treatment group attained the highest mean differ-ence scores on four measures: the Reading for Interpre-tation, Creative Comprehension, General'Comprehension, andComprehending Significant Ideas subtests. The LEA treat-ment group attained the highest mean difference score onthree measures: the Oral Reading, Reading for Relation-ships, and Comprehending Specific Instructions subtests.The LEA and Control Treatment groups attained the highermean:difference score for Literal Comprehension. TheControl treatment group attained the highest mean dif-ference score on four measures: the Vocabulary, AverageReading, Reading for Information, and Reading for Appre-ciation subtests.
Summary
This chapter has presented descriptive statisticsand results of analyses of covariance for posttestscores adjusted for the effects of the covariates. Meandifference scores were also presented. None of the dif-ferences among treatment groups were statistically sig-nificant.
30
Table 30
Pre -Posttest Difference Scores,by Treatment Groups
VariableTestLevel
NIM LEA ControlTreatment Treatment Treatment
RDTOral Reading
NDRTVocabulary
NDRTAverageReading
NDRTLiteralComprehension
NDRTReading forInformation
NDRTReading forRelationships
NDRTReading forInterpretation
NDRTReading forAppreciation
NDRTCreativeCompreP_msion
NDRTGeneralComprehension
NDRTComprehendingSignificantIdeas
NDRTComprehendingSpecificInstructions
PrimaryIntermedia
Primary &Intermediate
Primary &Intermediate
PrimaryIntermediate
Intermediate
Intermediate
Intermediate
Intermediate
Intermediate
Intermediate
Primary
Primary
1.0
.4
.6
.40.
3.4
. 2
.6
. 4
1.1
. 6
. 6
2.9
.6
1.3
.6
5
2.0
1.4
1.2
.2
.4
31
CHAPTER IV
CONCLUSIONS AND RECOMMENDATIGNS
Conclusons
Although growth in reading was evidenced in alltreatment groups, statistically significant differencesbetween the three groups were not revealed when the datawere submitted to analyses of covariance. None of thetwelve analyses revealed statistically significant dif-ferences. The null hypotheses of this study are there-fore not rejected. It can be concluded that neither theNIM nor the LEA nor the Control treatment was a more ef-fective instructional procedure than either of the others.
Discussion
The mean difference scores attained by subjects inthe NIM treatment in this study (.3-1.8) were lower thanthe gains reported by Heckelman (1.9-3.017). The meandifference scores attained by the LEA (-.2 to 1.1) andControl (.2 to 2.0) treatments in this study also weregenerally lower than the Baines reported by Heckelman.These data suggest that the Neurological impress Methodis an effective treatment procedure but may cast somedoubt on the original claims made by Heckelman.
This study-included subjects of different ages fromthe subjects in Heckelman's studies. The relatively high.(X s .40) correlation coefficients between age and read-ing achievement obtained in this study may have contri-buted to the differences in gain scores from this studyand from Heckelman's reports.
The other covariates used in this study might havehad an effect similar to that of chronological age inconfounding the problems of direct comparison betweenthe results of this study and the results reported byHeckelman. The lower limit of intelligence was 80 forthis study and 90 for the Heckelman studies. Achievementin spelling-was not considered by Heckelman but was in-cluded in--this study.
It is felt that the value of the NIM procedure asdemonstrated.by this .study, by Heckelman's studies, and
by Miller's study, is important. The writer is aware ofother case studies ,invviving NIM that have reported thisprocedure. as effective for remediating problems of dis-abled readers. It is 'believed that in light of the find-ings from research studies, gains in nonempirical casesshould not be ignored.
Limitations
Space restrictions in the. Purdue University Read-ing Clinic facilities limited the size of the sample._This small sample size may be a limitation of the studybecause statistical tests require much greater differ-ences among small groups than-among large groups toreach signifiCance.
The $15.00 application fee and $25.00 remediationfee required from parents of clients. in the Purdue Uni-versity Reading Clinic limited this study to subjectsfrom socioeconomic status groups able to pay-the fees.This restriction in the sample limits the population towhich results of this study may be generalized.
The differences in regularity of attendance among, Reading Clinic clients -were an uncontrolled lactor-
the study. Students receiving less instruction becauseof less attendance may have.depressed the mean scoresfor the groups.
Neither. time nor methods of reading instruction atschool are known-for the subjects in the experimentaltreatments. This lack of control over.reading instruc-tion provided outside-of the Reading Clinic may alsolimitthe credibility of treatment effects for this study.
Practice effects of repeating the_same -form of oraland silent reading tests may have'contributed to error ofmeasurement-within the study.
Educational Implications
The primary implication of this studristhat forbled .readers in general, there probably no one
best method_of instruction . This implication is. con-sistent with results of many studies involving typicalschool populations.
A further implication is that a.-teaching method ap-parently very effective for slected
not be similarly helpful for a larger group .of children.
Recommendations
A recommendation for further research is that treat-ment-be based on-differential diagnosis. Such factors aslearning modalities and personality characteristics maybe impol.tant determinant$ o4 treatment procedure.
A second recommendation for future research is thatattempts be made to control the level of teaching-profi-ciency of tutors. It is suspected that tutors having com-pleted a semester of remedial teaching in a clinical set-ting will be more effective than teachers without clini-cal experience.
It is recommended that future researchers controlfor the ages of disabled readers. In this study the rela-tively high correlation of chronological age with readingachievement was an unanticipated outcome.
34
REFERENCES
Arnold, Richard D. Reliability of Test Scores for theYoung'Bilingual Disadvantaged.' The Readiu
_
T.acher, 22:3413345, 969.
Bond, Guy L, Balow, Bruce, and Hoyt, Cyril J. New De-velopmental Readin Tests Intermediate Levi .--Chicago -Lyons an CarnahifT7-17976-87---
Bond, Guy L., Balow, Bruce, and Hoyt, Cyril J. New De-velopmental Readin Tests Primary Level:. Mica o:Lyons -iffa-Carna an, 19-65.
Bond, Guy L., Cir)or, Theodore, and Hoyt, Cyril J. De=veld mental Reading Tests Primary Level. Chicago;Lyons- art-. Carna an, 1961.
Bond, Guy L. and Dykstra, R. The Cooperative ResearchProgram in First Grade Reading. Reading ResearchQuarterly, Summer,.1967, p. 120.
Buros, Oscar Krisen (ed.) The Third Mental Measurement_Yearbook. New Brun vrEE:ErtaFfsMixrersity-Fi-JK7-1949.
Buros, Oscar Krisen (ed.) The Fourth Mental MeasurementsYearbook.- Highland P5 k, -Jersey: TEE-GTTAVEFF-67IT-Ig53.
Buros, Oscar Krisen (ed.) The Fifth Mental Measurements,Yearbook. Highland a- 1 JEFTET:-75E-GF7fffiTnress, 1959.
Buros, Oscar Krisen (ed.) The Sixth Mental MeasurementsYearbook. Highland Park,-NOW Jersey: Thd'Gryphon
1965.Pressg_
Buros, Oscar Krisen (ed.) The Seventh Mental MeasurementsYearbook. Highland Pk, New-UefFETT- TEFI7iTEEPie-FT:7972.
Campbell, Donald T. and Stanley, Julian C. Experimentaland Quasi-Experimental Designs for Research. Hand-book of Research on Teachini. Nathaniel Lees cage
icago RSEd7gERYTTy , -1963.
Coleman, H. A. The Relationship of Socio-Economic Statusto the Performance of Junior High School Students.Journal of Experimental .Education, 9:61-63, 1940.
Dixon, W. F. (ed.) Biomedical Computer Programs. Berkeley:University of California Press, 1970.
Dunn, Lloyd M. Peabody Picture Vocabulary Test. CirclePines, Minneota 7NriefiCan GUI-dant6 SW-Vice, Inc.1959.
Elashoff, J. D. Analysis of Covariance: A Delicate Instru-ment. American Educational Research Journal, Vol. 4,No. 3 , PP.
Fernald, Grace M. Remedial Techniques in Basic SchoolSub'ects. NeW-Yotk: ME-:467aT4--11111,
Gates, Arthur Irving. Gates Readin Diagnostic Tests.New York: Teachers761 ge u eau. Of PublIEFrons1945.
Gates, Arthur Irving and McKillop, Anne S. Gates-McKillo- Reading Diagnostic Tests. New Yerk:ea ers e Vre F71762.
Grannis, Chandler B. The Right and Ability to Read.Publisher's Weekly, 196:40, October 13, 1969.
Hahn, H. T. Three Approaches to Beginning Reading Instruc-tion - ITA0- Language Arts, and Basic Readers.- InThe First.Grade Reading Studies: Findings of Indi-VM5ITHVCiVriations. R71577tauffer (ed.T7Newark, Detawarei he International Reading Asso-ciation, 1967.
Harris, Albert J. How to Increase Readin Abilit_ (FourthEdition). Ne o aThrTiaThe ay mpany- Inc.,1961.
Harris, Albert How to Increase Readin Fi hEdition). NeTIYefF:=UETaMc ay-_ompany, nc.,1970.
Harris, Albert J. and Ser,er, B. L. Comparing ReadingApproaches in First. Grade Teaching with Disadvan-taged Children. In The First Grade Studies: Find-ings of Individual IFTZslia57 R777-5tauTIFF,
REWYFE7=1-11713TT7Thernational Read-ing Association, 1967.
36
Heckelman, R. G. 'A Neurological Impress-Method of Re-medial-Reading Iastruction. Academic Thera--Quarterly, Vol. IV, No. 4 :277 -282, Summer, _.69.
Heckelman, R. C. Using the Neurological Impress RemedialReading Techniques,'. Academic The a Quarterly,Vol. I, No. 4:230-239, SUmMer,
Hollingsworth, Paul M. An Experiment with the ImpressMethod of Teaching Reading. The Readin Teacher,Vol. XXIV, No. 2:112-114,187,Navem er, 1
Jastak, J. F., Bijou, S. W., and Jastak, S. R. WideRan e Achievement Test. Wilmington, Delaware:
anFiXTTc7EYFTes, 1965.
Johnson, M. S. and Kress, Roy A. Informal Reading Inven-tories. Newark, New Jersey:, internatienal Rila7ing-10-sociation, 1965.
Kaluger, George, and Kolson, Clifford J. ReadirrandLearnin- Disabilities. Columbus: C ar es ENerrill paumErHi-u7a. 1969.
Lee, Dorris M. and Allen, R. V. Lea nin to Read throughEx erience. (Second Edition Yirk: -App_etbn-e tury- rafts, 1963.
Miller, Marion Z. Remediation by Neurological Impress.Academic Therapy Quarterly, Vol. 4, No. 4, Summer,
Raven, J. C. The Coloured Progressive Matrices. London:H. K. Lew i and
Raven, J. C. Standard Progressive Matrices. London:LeWTSNfiffCo., 1960.
Reid, H. C.and Beitramo, L. Teaching Reading to theLow Group in First Grade. The Reading Teacher,19:601-604, May, 1966.
Robinson,-H. A. and Hanson,_E. Reliability of Measuresof Reading Achievement. The Reading Teacher 21:307-1130123, 1968.
Spieth, Phillip. A Contued Experimental Summer Programin Remedial Reading. Minnesota Readin Quarterly,Vol. XII, No. 3:66-73,IU4, February, 68.
37
Spieth, Phillip. A Second Continued. Experimental SummerProgram in Remedial Reading. Minnesota Readin-uarte Vol. XIII, No. 3:98A:042l: -e.ruary,
Spieth, Phillip. An Experimental Summer Program in Re-medial Reading. Minnesota Readin uaVol. XI, No. 3:817787riFfuary,
Veldman, D. J. Fortran Programming for the BehavioralSciences. NUITTY5rYT--HFIETRineEaif an Winston,167.
Weiner, B. J. Statistical Principles inDesi n OFFEFTUITIon New Yo 1oo ompany, 1971.
38
APPENDIX A
QUESTIONNAIRE
39
Dear
Your child, even though he has not been in our clinic thispast semester, has-been involved in a research study. We'are trying to determine if treatment at our clinic isworthwhile.
We need the following information about your child lastsemester:
A) To the best of your knowledge, did your child receivereadng instruction oth;r than he e u:arly receivesin his school classroom?
Yes
If you answered yes, please answer the following ques-tions.
1. Did he receive private tutorin
Yes No
How much time did this involve?week, 1/2 hour per day)
hour per
Did he receive any other special. help in reading?Yes No If so, please describe below andindicate the amount of time involved?
1 hour per
A copy of this questionnaire is also being sent to yourchild's teacher, who maybe helpful in giving us more"teacher-type information. The teacher will probablyfill in this section.
3. Did your child receive remedial reading at achool,Yes No How much time was involved?
hour FEF-
Did he receive any special help in reading atschool? Yes 'Nc How much time was in-volved?- our per
Your prompt and honest response is urgently needed.
Thank you for your cooperation.
Yours truly,
Richard D. AritoldAssociate Professor
Education
40
APPENDIX B
PERMISSION FORM
41
One of the primary purposes of the Remedial Reading Clinicof the Purdue Achievement Center for Children is to provide training and experience to students. Their trainingis advanced by working with and observing children.. In ad.-dition, the Center works closely with other professionalagencies in order to provide the most effective diagnostic,remedia., and consultative services. In order for us tofulfill our responsibilities both to students and to Otherprofessional agencies, we request that you read the follow-ing statement and sign your name below:
* *
We hereby permit the faculty, staff, and students of thePurdue Achievement Center for Children to test, to observe,to provide remedial assistance, and in other ways workwith our child; to video-tape-and audio-tape any of theabove activities as well as parent and.case conferences.
Also, we hereby permit t ePurdue Achievement _Center forChildren to acquire from and/or release to other professsional agencies (schools, clinics, doctors, etc.). any andall records and information pertaining to our child.
42