DOCUMENT RESUME

ED 058 833 HE 002 762

AUTHOR Walizer, Michael H.; Herriott, Robert E.TITLE The Impact of College on Students' Competence to

Function in a Learning Society.INSTITUTION American Coll. Testing Program, Iowa City, Iowa.

Research and Development Div.PUB DATE Dec 71NOTE 47p.AVAILABLE FROM Research and Development Division, The American

College Testing Program, P.O. Box 168, Iowa City,

Iowa 52240 ($1.00)

EDRS PRICE MF-$0.65 HC-$3.29DESCRIPTORS *College Graduates; *College Studenits; *Higher

Education; *Interpersonal Competence; *SocialDevelopment

ABSTRACTA considerable amount of evidence shows that college

has an impact on the personality characteristics of students in waysthat could be interpreted as giving competence to college graduatesto function successfully in leadership roles within our society.College graduates are more autonomous, independent, flexible, andsocially involved, for example, than noncollege graduates. Thispersonality difference could be due to student differences uponcollege entrance, or could be a result of the effects of thestructural aspects of colleges and universities as socialinstitutions. These 2 possibilities were explored by using the SocialMaturity Scale of the Omnibus Personality Inventory, and results showthat student change is related to the social structures of thecolleges attended as well as to student characteristics upon college

entrance. (HS)

U S DEPARTMENT OF HEALTH,EDUCATION & WELFAREOFFICE OF EDUCATION

THIS DOCUMENT HAS BEEN REPRODUCED EXACTLY AS RECEIVEO FROMTHE PERSON OR ORGANIZATION ORIGINATING IT POINTS OF VIEW OR OPINIONS STATED DO NOT NECESSARILYREPRESENT OFFICIAL OFFICE OF EDUCATION POSITION OR POLICY

PERMISSION TO REPRODUCE THIS COPYRIGHTED MA TERIAL HAS BEEN GRANTEDBY

e-.2_TO ERIC AND ORGANIZA

TIONS OPER NGUNDER AGREEMENTSWITH THE U S OFFICEOF EDUCATION FURTHER REPRODUCTION

OUTSIDI THE ERIC SYSTEM REOUIRES PERMISSION OF THE COPYRIGHT OWNER

THE IMPACT OF COLLEGE ON STUDENTS' COMPETENCE

TO FUNCTION IN A LEARNING SOCIETY

Prepared by the Research and Development DivisionThe American College Testing Program

© 1971 by The American College Testing Program

All rights reserved. Printed in the United States of America

For additional copies write:Research and Development DivisionThe American College Testing ProgramP.O. Box 168, lowa City, Iowa 52240

(Check or money order must accompany request.) Price: $1.00

ABSTRACT

Modern society, marked by a high level of technology and change, is aptly cal:A the"learning society" because people continually have to learn new roles. Those who function asleaders and managers of this society, the "technical stratum," go through a socializationprocess in which higher education plays an important oart. A considerable amount of evidencehas accumulated that college has an impact on the personality characteristics of students inways that could be interpreted as giving competence to college graduates to functionsuccessfully in leadership roles within such a society. College graduates are more autonomous,independent, flexible, and socially involved, for example, than noncollege graduates. Thischange in the personality characteristics of college students could be due to student differencesupon college entrance, or could be the effects of structural aspects of colleges and universitiesas social institutions. To explore these two possibilities, data were used from Trent andMedsker's Beyond High School study. The student outcome variable chosen to representsocialization to competence was the Social Maturity Scale of the Omnibus PersonalityI nventory.

Student change on the Social Maturity Scale was found to be related to the social structureof the colleges they attended as well as to student characteristics upon college entrance.Further, the social structure of the colleges by itself exerted a significant influence on studentSocial Maturity. When the data were reanalyzed controlling for the length of time studentswere in college and the initial scores on the Social Maturity Scale, the relationships wereenhanced. Limitations of the study are described, and implications for comparable researchwith other social institutions are suggested.

4

THE IMPACT OF P:OLLEGE ON STUDENTS' COMPETENCETO FUNCTION IN A LEARNING SOCIETY

Michael H. WalizerRobert E. Herriott'

Although in recent years almost all Americanyouth have become exposed to elementary andsecondary schooling as a form of socialization,today an expanding segment of American society ismaking its initial contact with higher education. Ineddition to simply making greater quantitativedemands upon the institution of formal education,such a rapid expansion in the "clientele" ofAmerican higher education has raised importantquestions regarding the functions of colleges anduniversities in contemporary society (Jencks &Riesman, 1968; Goslin, 1965; and Mayhew, 1969).Although a careful analysis of these many andoften competing functions would be useful, weshall confine our analysis to simply thesocialization function.

Contemporary America is experiencing socialchange at an unprecedented rate; the scientificrevolution, rapid urbanization, and phenomenalpopulation growth have all created a socioculturalcontext in which social change is, and will continueto be swift (Mack, 1967). One of the consequencesof such rapid change has been a shift from a"performance" to a "learning" society (Moore &Anderson, 1969), with a concomitant need forindividuals who can adjust and adapt to changingrole demands.

In a performance society one could be

reasonably certain that if he acquired the skillsnecessary to adequately perform his ascribed andachieved roles there would be a limited need :orsubsequent learning of new roles and modificationof established ones. However, in a learning societyan individual can no longer be certain that theroles, particularly occupational roles, he initiallyprepares for will adequately see him through hislife cycle. On the contrary, individuals can bereasonably certain they will be called upon tofulfill new roles and that the expectationsassociated with continuing roles will be modified asthe society changes. The ability to adopt andadequately carry out new and changing roles as thesociety changes can thus be viewed as a pre-condition to competence in a "learning" society.

The learning of such new or changing roles is aform of socialization. Although this concept hasbeen used predominantly in the study of childdevelopment, it can be viewed more broadly as"the process by which one learns to perform hisvarious roles adequately (Brim, 19681." More

1Th Is study was ,xomp feted In lows City during the summer of 1970as a pot of the work of the ACT Research Institute. Wa Ilzer is ananIslant professor of sociology at Western Michigan University andHerriott se professor of sociology at Florida State University.

5

1

specifically socialization can be defined as "anydevelopment which entails the modification of theself through the acquisition of personalitycharacteristics [Gottlieb, 1961] ." This process ofachieving effective participation in a social settingcan be called "socialization to competence."

Within every social system there are clusters ofpersonnel with a specific interest in influencing thebehavior and values of segments of the population,especially those of children and youth. Theseclusters can be viewed as sccialization agents ifthey have a program of socialization and possess asystem of norms, patterns of relationships, orstructures of roles (Parsons, 1951). In our societyhigher education functions, among other things, asa socialization agent consisting of "schools andcolleges, their academic practices, policies, andfacilities, and the structure of faculties and stu-dents who teach and learn in them [Sirjamaki,1967] ."

Two distinct approaches can be used in con-sidering the socialization function of highereducation in contemporary American society. Oneemphasizes what Mannheim (1940) has referred toas "substantial rationality" and focuses upon theperceptions of those engaged in the educationalprocess. Illustrative of such an approach is thework of Palley (1968) who performed a contentanalysis of the statements of objectives from thecatalogs of a large sample of colleges andu n i vers i t i es . However, a second approach

emphasizes "functional rationality" and focusesupon a series of deductions from known conditionswithin the larger society. A study by Hodgkins andHerriott (1970) proviee:, an example. Beginningwith explicit assumptions about the nature ofmodern American society, they examined thedevelopment of certain forms of social !earning notgenerally acknowledged by educators to be a part

of elementary schooling.

It is this latter approach which we shall follow.What we have described as a learning society can beviewed more generally as a society exhibiting ahigh degree of "modernity." Although this conceptis often the subject of disagreement (Anderson,1966), a consistent theme in all discussions of themodernization process is a society's dependenceupon the institution of formal education for thepreparation of personnel with both the cognitive

and social skills essential to the exploitation of a

2

rapidly advancing technology. Central among theconditions within a modern society which com-petent individuals must be able to confrontsuccessfully are increasing (a) urbanization, (b)specialization, (c) differentiation of the socialstructure (Durkheim, 1933), and (d) emphasis onfunctional authority (Moore, 1965). Thus, one ofthe functions of higher education in a modernsociety is the socialization of students to developthe personality characteristics enabling them toperform successfully in a highly urbanized,specialized, and differentiated social structure.(For a more extended discumion of the function offormal education in a modern society, see Herriottand Hodgkins, in press.)

One of the most explicit statements regardingthe specific personality characteristics required bysuch a society has been offered by Parsons andPlatt (in press). Collegiate socialization, consistentwith the principal value pattern of cognitiverationality, is seen to develop within the individualthe capacity to accept new levels of achievementfor self ana for others. This is closely articulatedwith the acceptance of necessary funztionalauthority connected with these levels of achieve-ment. Also, such socialization develops the,:apacity to participate in, and to accept, anextensively differentiated environment.

Broadly speaking, the "ideal product" of such aprocess of socialization can be seen as one who (a)tends toward universalistic standards of evaluation,(b) has the ability to evaluate and adjust to hisposition within a highly differentiated network,and (c) within this wide spectrum of differentia-tion and pluralism will accommodate necessaryauthority. In addition, the qualities of flexibility,autonomy, and ego strength are congruent withthese socialization goals.

These qualities would enable an individual toadopt and adequately fill a rather broad spectrumof roles within a modern aociety. In addition, theirexistence would increase the probability that hewould be predisposed to adapt to new roiedemands, have the capabilities to independentlyassess his position in a highly pluralistic society,and to achieve adequate role performance withsuticient personal adjustment. Thus, the socializa-tion function of higher education :n a modern,learning society can increasingly be seen as one ofsocialization to competence.

Socialization, Social Structure, and Competence

Inkeles (1966a) has proposed an approach forinvestigating the relationship between socialstructure and socialization to competence. I nkelesdefines competence as

the ability effectively to attain and perform in three sets of statuses:those which one's society will normally assign one, those in therepertoire of Ithel social system one may appropriately aspire to,and those which one might reasonably invent or elaborate foroneself.... (it includes) an individual's capacity to move to newstatuses and to elaborate new roles .... In our society this means,above all, the ability to work at gainful and reasooably remunerativeemployment, to meet the competition of those who would undo uswhile yet observing the rules for such competition set down bysociety, Zn manage one's own affairs, to achieve some significant andeffective participation in community and political life, and toestablish and maintain a reasonably stable home and family life (pp.265, 2801.

Such a notion of competence emphasizes theresult of previous socialization, but the concepts ofcompetence and socialization are linked throughthe social structure.

Socialization

The bulk of research on socialization has beenconcerned with the effects of familial experienceson individual development. Such phenomena aschild-rearing practices, family structure, parentalvalues, home environment, etc., have been used toexamine childhood socialization (Gottlieb, 1961;Inkeles, 1966a; Shibutani, 1961). In a recentreview, Clausen (1968b) observes that althoughsocialization is increasingly being used to denoterole learning at any age, it traditionally has been ageneric concept that

embraces child-rearing, education, enculturation, role learning,occupational preparation, preparation for marriage and parenthoodand, indeed, all social learning that is relevant to one's groupmemh3rships and life transitions 1p. 141.

This pervasive use of the concept has led Clausenand others to note that socialization has been usedto refer to an area of interest more than to anysharply defined process. These practices make thespecification of clear conceptual boundaries for theterm difficult. However, a distinction can be madebetween socialization and maturation ifsocialization is seen to entail learning and

3

maturation as the "unfolding of the potentialitiesof the organism which occurs more or less auto-matically except in the face of marked deprivatior[Clausen, 1968a, p. 5) ."

Our investigation will rely upon lnkeles for thesubstance of the concept of socalization. Thus,socialization will refer to the process whereby

individuals acquire the personal system propertiesthe knowledge,skills, attitudes, values, needs and motivations, cognitive, affectiveand conative patternswhich shape their adaptation to the physicaland sociocultural setting in which they live (Inkeles, 1969a, p.6151.

The Personal and Social Systems

In speaking of the personal system, I nkeles andLevinson (1963) refe i. to the totality of the stableattributes which characterize an individual. Thenotion of a "social system" connotes an emphasison interaction and interdependence of social

phenomena. By means of such an approach,separate social facts can be studied as parts ofwholes, and individual actors and actions in termsof patterns of interaction. Parsons (1951) bringsthese thoughts together when he defines a socialsystem as "a mode of organization of actionelemPnts relative to the persistence of orderedprocesses of change of the interactive patterns of a

plurality of individual actors [p. 241."Inkeles has presented a rather elaborate

accounting scheme for the personal system. Hisapproach is explicitly proposed as a practicalmodel of personality for use where discussion iscentered on interaction between the penonal andsocial systems. The elements of the personalsystem as seen by Inkeles are reviewed in Table 1.

Unfortunately, a formal accounting scheme forthe social system which would parallel that ofInkeles for the personal system is not available.However, there are analytical schemes which canbe used to study specific social systems. Forexample, lnkeles and Levinson (1963) present afour-part schema for the systemic study of anorganization. The schema includes: (a) ecologicalproperties such as physical characteristics,resources, and size; (b) cultural properties such astraditions, values, and goals; (c) structuralproperties which include the division of labor and

TABLE 1

An Accounting Scheme for Personality Studya

Psychomotor System TemperamentAptitudesSkills

Idea System

Motivational System

Relational System

Self-System

InformationOpinions and Attitudes

ValuesMotives and Needs

Orientation to Authority FiguresOrientation to Intimates and PeersOrientation to Collectivities

Conceptions of SelfModes of DefenseModes of Moral Functioning

Modes of Functioning Cognitive ModesAffective ModesConative Modes

aSource: Inkeles, 1966a, p. 267.

authority structure; and, finally, (d) social processcharacteristics, such as the actual workings of theorganization, the degree of formality, and theemotional climate. Other dimensions of a socialsystem can also be elaborated, includnig economicand political characteristics, systems of stratifica-tion, modes of social control, patterns of deviance,institutional complexes, the nature of role

prescriptions, distribution of authority, thesanction system, and the degree of consensus andconflict concerning goals (Rosen & Bates, 1967).

The basic assumption in Inkeles' approach isthat personality characteristics have an importantinflu,ence upon the acceptance and performance ofvarious roles, and that this influence is regular andsystematic. In the most general sense roles can beviewed as consisting of systems of expectationswhich exist in the social world surrounding theoccupant of a position (expectations regardingbehavior toward peopie in other positions andexpectations a person in a given position perceivesas applicable to his own behavior) and the specificbehavior of a person in a position when he

4

interacts with someone of another position(Deutsch & Krauss, 1965).

Traditionally, sociologists have explained satis-factory role performance as a function of thesystem of sanctions and rewards that are at thedisposal of society in order that the expectationsof society are fulfilled. lnkeles' approach, however,is more psychological. He asserts that personalitycharacteristics play an important part in recruit-ment to roles and performance in them, sinceindividual characteristics predispose persons tofavor one or another role and also affect thequality of role performance.

Social Structure and Socialization to Competence

The central fact which ties the personal andsocial systems together is that effective socializa-tion is a precondition for organized social life(Inkeles, 1969a). The sociologist interested insocial structure attempts to define, clarify, andexplain the causes and consequences of sets ofinstitutional arrangements. Since all patterns ofsocial organization are made up by actions ofindividuals, the personal system becomes the majori ntermedi at e mechanism between differentsegments of social structure.

Following the suggested approach to the studyof socialization to competence proposed by Inkeles(1966a) for a particular stratum of society, thisstudy will explore socialization to competence forpeople with a college education, hereafter referredto as the "technical" stratum. The approach will beto: (a) list demands required by a modern societyon individuals within the technical stratum, (b)specify the requisite personality qualities thesedemands require, and (c) investigate the socializa-tion patterns vis-avis the structure of highereducation which are presumed to promote these"socially demanded" personality dispositions.

Typically, a college graduate could expect to filloccupational roles ranging from managers andproprietors to executives and professionals. In

these various occupations the substantive concernof the individual may vary but generally there aremany similarities. The occupational roles of collegegraduates seem to be at a level involving adminis-tration, complex problem solving and decisionmaking where creativity, innovation, and personaladjustment are necessary and highly valued.

8

Requisite Personality Characteristics

We have briefly alluded to personality character-istics which might facilitate competent roleperformance in these occupational roles. However,further explication of these personality character-istics is in order.

Much speculation has occurred concerning thepersonality characteristics required of individualswithin the technical stratum of a modern,"learning " society. For example, Lerner (1958)points to the importance of the need for empathyand in describing the "mobile person" he notesthat both independence from traditional patternsand autonomy of thought are necessary forcompetence.

Similarly, Lundberg (1963) discusses what hecalls the "man of thought." In a complicatedsociety the ability to make decisions in a complexcontext is prerequisite for maximum personaldevelopment. Lundberg also states that the aware-ness of an individual to personal and occupationalalternatives is more likely if the person is flexibleand open.

Inkeles (1966b, 1969b) empirically attempts toverify the existence of personality characteristicswhich would typify "modern man." The develop-ment of personality characteristics typical ofmodern man is seen to be a matter of degree, andas such technical stratum individuals can be viewedas having these characteristics to a higher degree.

Such characteristics include: (a) an openness to newexperience, (b) independence from traditional andhierarchical authority, and (c) full participation inone's social world.

The importance of autonomy in achieving one'spotential has been noted by Jahoda (1959),Maslow (1962), and Rogers (1951). Autonomyincludes the notions of "flexible, objectivethinking and an openness of attitude whichfacilitates awareness of and adaptability to theenvironment [Trent & Medsker, 19681 ." Creativityhas been intimately related to autonomy in theworks of Barron (1961), Crutchfield (1963),Gough (1961), Nelson (1961), MacKinnon (1961),and Sanford (1966).

As Trent and Medsker (1968) observed,autonomy has been related to intellectual aware-ness in a manner similar to that which has beenused in relating an authoritarian personality to

nonintellectuality. Trent and Medsker (1968) con-clude,

the autonomous individual is capable of the objective, open, andflexible thinking which characterize intellectuality, and theauthoritarian individual is distinguished by the highly opinionated,clorai thinking which is the mark of nonintellectuality [p. 111.

In summary the personality characteristicsnecessary for competence in technical stratumoccupational roles of a learning society wouldinclude:

tolerance for ambiguitiescreativityan open, receptive mindcritical thinking abilityflexibilityfreedom from authoritarianism and opinionated

thinkingrecognition of achievement via independenceability to think abstractly and reflectivelyability to assimilate new information in a logical

manner.

Effects of Higher Education as Socialization toCompetence

The bulk of evidence currently available in theliterature concerning the effects of collegeindicates that students do change during thecollege experience. Much of the research has beendirected at specific substantive areas of studentattitudes, opinions, beliefs, or behavior; but anotable portion deals with a more general level ofpersonality change. Those studies which do dealwith the personality characteristics outlinedpreviously may be viewed as dealing with thosepersonal system characteristics which can beassociated with competence.

Perhaps a major impetus for much contem-porary concern with the effects of college

education was the study done by Jacob (1956) andthe subsequent discussions by Riesman (1958) andBarton (1959).

Comprehensive reviews of the literature oncollege effects done by Freedman (1960),Lehmann and Dressel (1963), Newcomb andFeldman (1968), and Webster, Freedman, andHeist (1962) exhibit a consensus that college

5 9

student personality characteristics do change.

Attitudes toward public issues become more liberaland a greater sensitivity and appreciation ofaesthetic activity develops (Berdie, 1968; Eddy,1959; Elton, 1969; Elton & Rose, 1968; Feldman& Newcomb, '1969; Hodgkins, 1964; and Pace,1941).

One of the most recent and perhaps mostcomprehensive studies of changes occurring at theindividual level as a result of college attendance isthat of Trent and Medsker (1968). In a longitu-dinal study beginning with over 10,000 high schoolgraduates, the investigators found that generally

speaking during college, students become morecritical in their thinking, more tolerant, moreflexible, and less prejudiced in their judgments.These findings are supported by the work of manyresearchers, particularly: Allwood (1964),Freedman (1960, 1967), Goldsen ( 1960), Katz et al.(1968), Lehmann and Payne (1963), Newcomb(1943), Newcomb et al. (1967), Webster (1958).

The evidence seems conclusive that the collegeexperience develops in the technical stratum ofindividuals those personal system characteristicsindicative of what we have identified as com-petence in a modern, learning society.

Approaches to the Study of College Students' Interaction with Their Environment

Most investigators of the collegiate experiencewould agree that changes in personality occur instudents during the college years. Whether thesechanges are best attributable to "maturation" or"socialization" is far more uncertain, but there issome evidence, particularly that of Trent andMedsker (1968), to the effect that it is the collegeexperience per se which makes the difference. Forthe time being, let us accept the premise that thecollege experience can produce changes inpersonality and consider what it is about .thatexperience which can produce such change.

In most past research personality changes of thetype described above have been assumed to resultfrom a process of interaction between the studentand elements of his collegiate "environment." Veryoften- such an assumption is not made explicit, asinvestigators frequently rely upon variables whichare simply suggestive of such interactions, ratherthan tapping the interactions directly. The typicalapproach has been to identify social-psychologicalaspects of the student's environment while incollege which are indicative of conditions that canfoster certain interactions. A second general

strategy used by past researchers has been theidentification of individual student characteristicswhich would predispose him to participate inspecific forms of interaction. These individualcharacteristics have then been used as proxyvariables for the interaction patterns thought mostlikely to produce personality change.

Formal studies which have attempted to explorethe impact of college on student personality usingthese approaches can be classified into threegeneral groups: (a) the "subcultural" approach, (b)the "subjective environment" approach, and (c)the "residual" approach. Although assignment ofany particular study to one of these groups may bedifficult due to overlap, generally studies seem tocorrespond to the categories listed.

The Subcultural Approach

By far the approach which has been used mostfrequently in studies of college effects is one whichemphasizes the student subculture. Stemming fromsuch notions as that of W. I. Thomas that "it is theculture of the group that limits the power of themind to meet Crisis and to adjust," contemporaryinvestigators have argued that

the groups to which a person belongs set the limits, provide thealternatives, and define the meanings to be attributed to threateningas well as nonthreatening situations. . . . it is a person's sociallyrelevant groups that train the individual for legitimate and propernodes of adaptatiols (Mechanic, 1962, p. 71.

For students in general and particularly for highschool students the peer group is thought to be theSocially relevant group. Coleman (1961), Gordon(1957), Havighurst and Taba (1949), Hollingshead(1949), and others reasoned that peer subcultureplays' a predominant role in students' lives. How-

ever, as Turner (1964) pointed out, the youthculture has been the subject of impressionisticobservation rather than controlled investigationand "the content of youth culture and kinds ofexisting hierarchies are still largely unknown."

The studies of student subcultures of collegestudents seem to follow the general pattern of,first, identifying dimensions of thought, attitude,or behavior along which students might vary, and,

second, arriving at various subcultures by develop-ing categories on these dimensions. Perhaps thebest known of these typologies was offered byClark (1962). He classified students along twodimensions: involvement with ideas and identifyingwith their college (see Figure 1).

Identify withTheir College

Involvement with Ideas

academic collegiate

nonconformistvocational-

consumer

Fig. 1. Typology of student subcultures. Source: Clark(1962), p. 210.

The "collegiate" subculture is characterized byfootball, fraternities, dates, and campus fun. The"vocational-consumer" subculture is almost solelyinterested in job training. Serious students whoidentify with intellectual concerns are placed in the"academic" subculture. And finally, the "noncon-formists" are seen as alienated, involved with ideasof the adult society and as using off-campus groupsas points of reference.

In a similar manner Hodgkins (1964) arriVed atfour narrative descriptions of subcultures whichwere differentiated on the basis of the emphasis

placed upon three general developmental aims ofhigher educationthe academic (intellectual), thesocial, and the vocational. Viewing the academicenvironment as a distinctive sociocultural systemGottlieb and Hodgkins (1963) demonstrated thatstudents who fit these varying nbrrative descrip-tions undergo different forms of attitude change.

1

T h ist I ethwaite and Wheeler (1966) haveemployed the idea bf student subculture in thearea of aspirations, while Goldsen (1960) utilizingthe notion of subgroups investigated issues rangingfrom goals of education to dating patterns.Riesman (1961) suggests that peer group sub-cultures play an important role in the process ofchange during the college experience. Newcomb(1961) has attempted to draw upon the variousstudies of student peer groups in providing ataxonomy for peer gtoup formation and conditionsof peer group influence.

Bolton and Kammeyer (1967) take a ratherstrict position when they define subculture as

a normative value system held by some group of persons who are inpersisting interaction, who transmit the norms and values tonewcomers by some communication,/ process and who exercisesome sort of social control to ensure conformity to the norms.Furthermore, the normetive system of such a group must differfrom the normative value system of the larger, dominant society [p.1251.

Utilizing this definition they suggest the conceptof subculture appears to be often misused in thestudy of college student peer groups. Alternativeconcepts such as "social type" (Klapp, 1962) and"role orientation" (Bolton & Kammeyer, 1967)have been used to denote the same generalphenomenon referred to by the term "subculture."

The Subjective Environment Approach

The second approach for specifying environ-mental factors for successful socia!ization con-stitutes what can be called the "subjective environ-ment" approach. These attempts at specifying thefunctional aspects of college environments ofstudents rely upon the students' subjective reportsof their college environments.

Among the most notable work conducted withthis type of orientation is that of Pace (1964) andStern (1963). Utilizing the notion of environ-mental press and congruence of this press withindividual needs, the "college characteristicsanalysis" was first developed. In conjunction withtheir work two other measures involving eratiron-mental press have been devisedthe ActivitiesIndex (Al) and the College Characteristics Index(CCI). In essence, the individual, through hiscollege experienr 7, describes in terms of environ-mental press his mental picture of his college

environment. Through this scheme Pace and Sternhave classified environmental emphasis and utilized"college press" as a major independent variable inarriving at a variety of conclusions. (See Stern,1963, for a summary of research findings.) Thiswork suggests there are two major factors whichaccount for most of the differences among collegeenvironmentsintellectual and social climates.

Another psychologically oriented attempt toclassify college students' environments was doneby Pervin (1967). Based on the semantic differ-ential, he has devised TAPETransactionalAnalysis of Personality and Environment. With thisinstrument various concepts concerning self andcollege can be rated. Pervin found considerablevariability in scale rating across the 20 collegesstudied on each concept and in the pattern ofratings across the six concepts of: my college,myself, students, faculty, administration, and idealcollege. However, to date, use of the instrumenthas been limited to studying satisfaction withcollege. In general, discrepancies between studentperceptions of themselves and of their college werefound to be related to dissatisfaction with college.

Eddy (1959) has shown that the student'sexpectancy of the institution can have an influenceon the impact of that college on student character.And Eckert (1943) has shown that pressure onstudents to be interested in current affairs pro-duced increased interest in current events.

Thistlethwaite 11959), utilizing the "college press"approach, has demonstrated that the collegeenvironment is an important determinant ofstudents' motivation to seek advanced intellectualtraining. Finally, in a review of many studies, Plant(1963) concludes that changes in attitudes dooccur among college students; and he summarizesthe impact of various phenomena.

The Residual Approach

The third category of attempts to define thefunctional aspects of college environments has nocommon element to bind the studies together.Some could be placed in the "subjective environ-ment" approach and some have elements of the"subcultural" approach. However, they do nothave the consistency of the previously mentionedstudies.

The work of Astin and Holland (1961) can serveas an example of attempts to assess subjectively thecollege as a student's environment. The Environ-mental Assessment Technique (EAT) is based onthe assumption that environments are transmittedby people. Further, since the college peer group isthe student's major personal contact, it followsthat a chief portion of the student's environment isdetermined by the characteristics of his fellowstudents; however, characteristics of the college asan institution are also included. Consequently, theEAT is defined in terms of: average intelligence ofthe students, size of the college, plus six personalorientationsrealistic, scientific, social, con-ventional, enterprising, and artistic.

In a related work Astin (1962), with a sample of335 institutions, performed a factor analysis of 33major attributes of colleges. Included in theseattributes were such factors as type of control,religious affiliation, faculty and student character-istics, and student "environment." The sixprincipal dimensions along which institutionsappeared to differ were identified as: affluence(wealth), size, type of control, masculinity/femininity, realistic (technical) emphasis, andhomogeneity. Among these six, affluenceaccounted for the largest proportion of the totalvariation.

Empirical investigations utilizing the EAT andthe latter factor analysis have shown some inter-esting results. Astin (1963) and Nichols (1964)have shown that colleges differ in their influenceon such student characteristics as achievement,ability patterns, and career choices; but the effectsof the college are small compared to differenceswhich are attributable to the characteristics of thestudents at the time of college admission (Nichols,1967). Nichols also found that college studentsbecome more aware of their own impulses and lessdependent. However, only about 5% of the totalvariation seems to be attributable to characteristicsof the college, such as the type of degrees whichare granted.

Astin's most recent major attempt to character-ize colleges as student environments has taken aslightly different approach. In what he himselfdescribes as a "shot-gun" method, placing anemphasis on behavior as opposed to perceptions (asin the CCI) or personal characteristics (as in theEAT), he has deVeloped an Inventory of College

12

Activities (ICA) which includes such items as theamount of time spent studying and the frequencyof intellectual arguments. Astin (1968) states that

our goal in designing the ICA was to identify as many environmental

stimuli as possible that could be observed by undergraduate studentsand reported in a questionnaire. We tried to cover four broadcategories of stimuli: the peer environment, the classroom environ-ment, the administrative environment, and the physical environment

IP. 91 .

Other scattered attempts at specifying theeducationally salient aspects of the college

student's environment range from the very vague,such as Jacob's (1956) fleeting reference to the"climate" at certain "special" colleges, to the veryspecific, such as Learned and Woods' (1938)finding that students who went to colleges wherethe average ability was high performed significantlybetter on comprehensive achievement tests thandid students of comparable ability who went tocolleges where average ability was low. Alsoincluded in this category might be such studies asNewcomb's Bennington College studies where thelocal community was found to be salient forchanges in political attitudes. Mechanic's (1962)viewing of the student's environment as a com-munication system and the consequent importanceof physical location and potential for interactionwith other students and faculty might also beclassified in the approach.

The Structural Approach

Noticeably lacking in the above studies is anexplicit emphasis on structural or organizationalcharacteristics of colleges and universities in

explaining the impact of varying institutionalenvironments on college students, although ourreview does show that the impact of college onstudents is generally viewed as socialization ratherthan merely as maturation. In addition the person-ality characteristics developed as a result of thecollege experience coincide with those specified asbeing requisite'to competence in a learning society.Because of the importance of structure to socializa-tion (as previously discussed) it seems appropriatethat greater attention be drawn to the structure ofinstitutions of higher education and its importanceto their impact on sttidents.

We know of no studies of the impact of collegeson students which have utilized .a comprehensive

strUctural approach. However, there are studieswhich have utilized structural variables in part oftheir analysis (Astin & Holland, 1961; Rogoff,1957).

In a recent study by Scott and EI-Essal (1969)student protest was studied in relationship tostructural and analytic variables. Size of school,quality of school, size of the community in whichthe institution is located, and the complexity ofthe institution were utilized. Similarly, Trent andMedsker (1968) utilize type of control in com-paring changes in "social maturity."

Parsons and Platt (1968) have characterizedacademic institutions by developing a scale ofdifferentiation. The Scale of Institutional Differ-entiation (SID) is intended as an index of thedegree of internal differentiation found in a par-ticular college as a social system. The scale iscomposed of three indices: size, quality, andresearch orientation (Parsons & Platt, 1968). Toarrive at size each full-time faculty member in thearts and sciences sectors of the institution wascounted as one, part-time as one-half, and profes-sional faculty was omitted. Quality was defined bythree sub-indices: general and educational incomeper student, the proportion of PhD holders on thefaculty, and the student/faculty ratio. Researchorientation was also derived from three sub-indices: the percentage of the arts and sciencesstudent body that are graduate students, themonetary amount of grants received by an institu-tion per faculty member, and the number ofperiodicals in the library per faculty member. TheSID appears to be a particularly promising devicefor measuring such structural phenomena as alloca-tion of academic role functions and teaching goals

and styles.Frequently academic "quality" has been utilized

in attempting to explicate the stratification systemof colleges as social systems. Traditionally aca-demic quality has been defined in either of twoways. One method defines quality in terms of anindex in which a set of indicators (generally oforganizational input) are assumed to adequatelyportray quality. The second method relies uponsubjective evaluations (generally of organizationaloutput) by judges who are considered authorities.

Knapp and Greenbaum (1953) assessed aca-demic quality of undergraduate colleges anduniversities on the basis of their production of

1 3

scholars and rated the institutions accordingly.Lazarsfeld and Thielens (1958) acknowledged thefutility of attempting a general definition ofacademic quality and resorted to operationallyassigning a quality rating to each college, utilizingsix indicators to measure quality. The most notableattempts of institutional rankings have beenperformed by Brown (1965) and Astin (1965).Astin employed 13 separate indicators in attempt-ing to arrive at meaningful assessment of studentenvironments associated with various colleges.Some of the resulting factors are utilized to create"a reasonably accurate measure of prestige."Brown utilized the following eight indicators toarrive at a final ranking of institutions into sixclasses:

1. Percentage of the total faculty with a PhD

2. Average salary and fringe benefits per facultymember

3. Percentage of students continuing on to grad-uate school

4. Percentage of students at the graduate level

5. Number of volumes in library per full-timestudent

6. Total number of full-time faculty members

7. Faculty-student ratio

8. Educational and general income per full-timestudent.

Although these illustrations demonstrate thatstructural features of institutions of higher educa-tion have been considered in past research, theyrepresent only a small minority of studies andsuggest that a thorough structural analysis of theimpact of higher education on the development ofcompetence in a learning society is likely to provefru itfu I.

Design of the Study

The goal of this report is to develop anaccounting scheme for the structural analysis ofthe social systems oi institutions of higher educa-tion which would parallel that presented by lnkelesfor the personality system. We will then examinechanges in the personality system reflecting com-petence in a learning society which can be attrib-uted to the influence of the social structure ofinstitutions of higher education.

An Accounting Scheme for College Structure

Ideally, a study of the relationship between thepersonal system and the structural properties ofhigher education would fully specify, utilizingInkeles' scheme of the personal system (see Table1), all of the personality attributes in all of thedimensions of the personal system which would berequisite to competence for the college-educatedstratum of society. In addition, a similar account-ing scheme for the structure of institutions, ofhigher education would be developed and each

10

salient dimension within that scheme would beexplicated and related theoretically to the personalsystem. Measurement devised for each accountingscheme would then be developed and administeredto a random sample of students in a randomsample of institutions of higher education using amatched longitudinal design. Extensive descriptiveand mu ltivariate analysis then could be carried outcontrolling for the previous socialization experi-ences of the students. For pragmatic reasons wehave had to adopt a more limited approach, butthe general procedures outlined will be used.

Table 2 illustrates our proposed accountingscheme for the study, of the structure of institu-tions of higher education. The left side of the tablepresents various classifications by which Structuralcharacteristics of colleges and universities can bedelimited and compared or contrasted. At the rightof the table are examples of the types of structuralrelationships which can be placed in the variouscategories. For example, the authority system ofcolleges and .universities can be compared with

14

TABLE 2

An Accounting Scheme for the Study of the Structureof Institutions of Higher Education

Structural dimension

Authority System

Example

Degree of BureaucratizationCollegial versus Authoritarian

Decision MakingLegitimacy of Power Relationships

Reward System Nature of RewardsMechanisms for Dispersement

Sanction System Nature of SanctionsMechanisms of Dispersal

Goal System Diversity of GoalsExistence of GoalsConsensus on Goal,

Personnel System Demographic CharacteristicsDegree of Training

Environmental Ex-change System

Type of ControlSource of FundingRelationship to College CommunityRelationship to Other Institutions

Ecological System SizeSurroundingsPhysical Facilities

nteraction System Time and Energy Allocationsfor Various Interactions

respect to the extent to which important decisionswithin the organization are made on a collegialbasis or through administrative authority. The goalsystem of the organization can be differentiatedon the basis of the diversity of goals (e.g., thenumber of degree granting departments), and/orthe consensus of goals within the organization.

Research Hypotheses

In order to conduct a preliminary test of theutility of such an accounting 'scheme to theexplanation of changes of personality duringcollege the following hypothesis is offered. Wepropose that personality changes eiperienced by

college students will be a function of the socialstructure of the colleges which they attend. We willcall this Hypothesis One.

However, there is a considerable body of pastresearch which suggests that personality changeduring college may simply be a function of theprecollegiate socialization to which college stu-dents have been exposed. Such socialization can beviewed as attributable largely to the social settingsof family, school, and church. Experiences withineach of these three alternative socializing agenciestheoretically can create preconditions for person-ality changes which will take place later, and canalso influence the types of college structures whichadolescents later encounter.

Table 3 presents a similar ace. nting scheme forthe structural analysis of precollegiate socializationbackground. Emphasizing the three social institu-tions just noted which can play a predominant rolein precollegiate socialization, the accountingscheme presents dimensions of each social institu-tion whIch can be used to characterize thatinstitution, as well as examples corresponding toeach dimension.

Since there are possible confounding effects toprecollegiate socialization with socialization incollege, it is important to test the assumption thatprecollegiate socialization experiences are relatedto personality change while in college. If thisassumption holds, we propose to further test thehypothesis that personality change experienced bycollege students will be a function of the collegesocial structure, after controlling for the possibleextraneous effects of precollegiate socializationexperiences. We will call this Hypothesis Two.

At this point in the development of our analyticframework it may be helpful to represent thesetwo hypotheses using several Venn-diagrams, ageometric fortp of set notation. Portrayed inFigure 2 are four alternative models of the possiblerelationship between college structure variables,precollegiate socialization background variables,and personality 'change. TIbe large outer circlesrepresent the total variance across a predefinedpopulation of students on some appropriate mea-sure of personality change (Vp). The inner circle,labeled Vb, represents the proportion of Vp whichis attributable to precollegiate socialization, andthe inner circle, Vp, the portion of Vp attributableto college structure.

TABLE 3

An Accounting Scheme for the Structural Analysis ofPrecollegiate Socialization Background

Socialinstitution

FAM I LY

SCHOOL

RELIGION

Structural dimension

Authority StructureKinship StructureValue StructureKnowledge StructureEconomic StructureEcological StructureAffective StructureReward & Sanction System

Authority SystemReward & Sanction SystemGoal StructurePersonnel SystemEnvironmental ExchangeEcological SystemInteraction System

Belief SystemValue SystemDifferentiation

of Activity

Example

Exercise of PowerExtended vs. NuclearAttitudes and BeliefsEducational AttainmentsSocioeconomic StatusSiblingsLove RelationshipsSocialization Practices

Degree of BureaucratizationNature of RewardsConsensus of GoalsTeacher PreparationSource of SupportStudent EnrollmentDaily Activities

Degree of MysticismAttitudesYouth Groups

Model I

Fig. 2. Four alternative models for the(Vb) and personality change (Vp).

Model II

Model IV

relationship of college strut.ure (Vs), precollegiate background

Notice in Figure 2 that in all four alternativemodels we have portrayed the "union" of Vb andVc (i.e., the area bounded by either Vb or V c) tobe less than V . Although the theoretical frame-work proposed in the present discussion introducesonly Vb and Vc as possible explanations of VD,factors outside of this framework also potentiallyinfluence personality change.

Hypothesis One specifically deals with the issueof the magnitude of Vc. It is proposed that Vc willbe non-zero.

These four alternative models are particularlyrelevant to an understanding of Hypothesis Two.Notice that in Model I, Vc and Vb do not intersect(i.e., overlap) suggesting a conception of precol-legiate socialization and college structure as statis-tically independent of each other. Such a model isoffered exclusively for illustrative purposes for it isnot very realistic given the extensive literaturewhich has shown that family background andchoice of college are often highly correlated (Astin,1965). Models II, I ii, and IV are more realistic andreflect an "intersection" between Vb and V. InModel II this overlap is rather minor, in Model IIImajor, and in Model IV total. Hypothesis Two isspecifically addressed to the issue of whether theappropriate model is Model IV or one more like IIor III. It predicts that the proportion of variationin personality change attributable to college struc-ture is not simply a subset of that attributable toprecollegiate socialization.

Ideal Research Design

In an ideal research design for testing these twohypotheses we would want to randomly selectstudents who were enrolled full time from ran-domly selected institutions of higher education. Asearly as possible in the student's college career thecompetence level of each student, as previouslydefined, would be determined. At this pointinformation from the respondents would becombined with data from familial, school, church,college, and other official sources to determine (a)the precollegiate socialization background of theindividuals with reference to those factors whichwould effect the further development of compe-tence and, (b) the social structure of the institu-tions of higher education. During the students'collegiate experience measures of the extent to

0 .

which they were exposed to the college or univer-sity as a socialization agent would be applied sothat the extent of exposure could be determined.Finally at the end of their college attendance,whatever the cause of termination, the competencelevel of each student would again be measured.

With this research design, a full test of theaccounting scheme which has previously beenpresented could be achieved.

Exploratory Research Design

Prior to a full test of the analytic frameworkdeveloped here, we conducted research of a prelim-inary nature. The ideal research design would becostly to implement in terms of personnel andfinancial resources, and due to the longitudinalnature of the suggested design a number of yearswould be required for the full test. However, auseful preliminary test of the hypotheses can becarried out utilizing data from an existing longi-tudinal study.

To conduct such a test of the hypotheses, asecondary analysis was performed on a portion ofthe data from the nationwide study Beyond HighSchool (Trent & Medsker, 1968). Through cooper-ation with the Center for Research and Develop-ment in Higher Education in Berkeley, California, aportion of the Trent-Medsker data was obtainedconcerning those 3,927 students from this studywho enrolled full time in higher education afterhigh school. This information included data on thestudents' background, including high school andcommunity information; longitudinally measuredpersonality characteristics; and information as tothe students' participation in higher education.

The Available Sample

The Trent-Medsker sample was drawn originallyto investigate relationships between college atten-dance and the type of college available in thecommunity. Community characteristics wareobtained using the Market Guide of the Editor andPublisher Company (1957) and documents of theU.S. Bureau of the Census (1952, 1957). Sixteencommunities were selected which were roughlycomparable (with the exception of one metropol-itan area for contrast purposes) according tothe following criteria:

1713

1. Population. For the 15 "basic" communities thepopulation range according to the 1950 censuswas 22,467 to 115,911, with a median of40,517. The population of the metropolitancommunity was 775,357.

2. Percentage engaged in white-collar occupations.The percentage of the population engaged inwhite-collar occupations ranged from 31 to 50%,with a median of 44%. In the metropolitancommunity, the percentage was 53.

3. Percentage of population engaged in manufac-turing. The range in the percentages of thepopulation engaged in manufacturing was 6.9 to56.2, with a median of 34.3. In the metropolitancommunity, the percentage of those engaged inmanufacturing was 16.9.

4. Percentage of the population engaged in trade.The percentage of the population engaged intrade, primarily in sales and retail, ranged from15.1 to 34.5, with a median of 21.6. In themetropolitan community, 25.5% of the popula-tion was engaged in trade.

5. Type of principal industries. The mediannumber of types of principal industries for the15 communities was 5, with a range of 4through 9. The one large metropolitan commu-nity had 8 such industries.

6. Median salary. The median salary for all thecommunities, based on the median within eachcity, was $3,335. The median salary in thecommunities ranged from $2,600 to $4,374,including the metropolitan community.

7. Median grade of education completed by those25 years or over. The median grade level ofeducation completed by those 25 years or overfor all of these communities combined was 10years; the range of grades completed for thisgroup among the communities was between 9.2and 11.5 years. In the case of the metropolitancommunity, the median grade was 11.7.

8. Ethnicity. The percentage of foreign-born white,both male and female, ranged from 1 to 15, witha median of 4. The metropolitan community

14

had the largest percentage of foreign-born white,namely 16%. In two communities no Negropopulation was reported. The range of the Negropopulation in the remaining communities ex-tended from 2 to 8%, with the median at 4%.Six percent of the population was reportedNegro in the metropolitan community (Trent &Medsker, 1968, pp. 273-274).

The communities and their characteristics aregiven in Table 4. The sample was composed of theentire seniur class of 1959 in public and privatesecondary schools (where private enrollment wasappreciable) in all of the communities except thelargest one (San Francisco). In the latter, seniorswere surveyed in three high schools representing ademographic cross section as determined by theschool superintendent.

Data concerning the students' high schools wereobtained from questionnaires to school principalsin the spring of 1959 and the fall of 1962. Data onhigh school achievement and measured intelligencewere gleaned from official records. Through surveyquestionnaires, data concerning student back-ground factors, school activities, and personalitycharacteristics were collected in the spring of 1959at the various high schools and again in the springof 1963. The follow-up instruments were adminis-tered by officials in the more than 50 collegeswhich enrolled 10 or more of the sample and inother cases response was obtained by mail. Addi-tional information was obtained from postcardquestions and from officials at the various institu-tions of higher education.

A total of 2,404 students who entered highereducaton full time in the fall of 1959 were studiedlongitudinally. A total of 683 institutions of highereducation were attended by these students. Amajority of the students, 69.2%, attended only onecollege during the 4-year period. The number ofsampled students at a given institution of highereducation varied from 1 to 252.

Research Variables

Dependent variable. Ideally in testing Hypothe-ses One and Two the dependent variable shouldencompass all of the elements of personality whichare descriptive of competence in a learning society.An explication of such characteristics, utilizing

18

TABLE 4

Characteristics of Participating Communities(Trent & Medsker, 1968, p. 275)

Community "rype of localcollege

Number ofparticipatinghigh schools

Popula-tion

Percentwhite-collar

percentmanufac-

turing

Per-

centtrade

Number of Median Mediantypes of salary gradeprincipal COrW

industries pleted

Percentforeign-

bornwhite

Percent

NNIro

Altoona, Pa. Extension center 2 77,177 35 11.0 21.5 8 $2,907 9.5 5 0Bakersfield, Cal. Junior college 3 34,784 33 6.9 27.6 4 4,374 11.5 6 4Danville, Ill. Junior college 1 37,863 44 24.6 24.6 5 3,292 9.2 2 8Eau Claire, Wis. State college 2 36,058 44 34.3 23.60 5 3,670 10.8 4 0Freeport, 111.8 None 1 22,467 46 25.00 -- 3 3,064 9.4 4 3Hutchinson, Kan. Junior college 1 33,575 47 16.5 30.9 7 3,147 11.0 2 3Joplin, Mo. Junior college 1 38,711 50 16.0 34.5 5 2,600 9.9 1 2Kalamazoo, Mich. State coliege 4 57,704 44 39.1 19.8 4 3,593 10.7 7 4Loraine, Ohio None 1 51,202 31 56.2 15.1 4 3,681 9.5 15 5Muncie, Ind. State college 3 58,479 37 44.4 20.3 4 3,335 10.0 1 8Port Huron, Mich.b Junior college 1 35,725 44 34.7 21.6 4 3.472 10.1 10 3Racine, Wis. Extension center 4 71,193 39 55.2 16.7 5 4,051 9.5 12 4San Francisco, Cal. Variety of colleges 3 775,357 53 16.9 25.5 3,923 11.7 16 6South Bend, Ind. Extension center 5 115,911 47 50.8 18.0 5 4,349 10.1 8 7Springfield, Mo. State college 3 66,731 49 13.9 30.4 9 2,819 10.6 1 3Zanesville, Ohio Extension center 2 40,517 44 40.2 22.8 4 3,064 9.2 2

allo information could be found regarding the percentage of the community population engeged in trade; manufacturing was determined on acountry-wide basis.b Information on types of principal industries was available for this community only on the basis of classification uf industries employed by TheEditor and Publisher Market Guide (1957).

Inkeles' accounting scheme as a guide, has beendiscussed.

The dependent variable for the present test ofthe proposed hypotheses is much more modest andfocuses on the development of social skills asmeasured by the Social Maturity (SM) scale of theOmnibus Personality Inventory (OPI)? The OPIwas developed at the Center for the Study ofHigher Education at the University of California(Berkeley) to measure certain areas of ego-functioning for the purpose of providing

a meaningful, differentiating description of students and a means ofassessing change rIther than a device or instrument for testing aspecific theory of personality. n. approach to assessment ofhuman behavior through several related domains was planned anddeveloped as a meaningful way of studying students in a variety oflearning contexts (Heist & Yonge, 1968, p.

In spite of its name the scale measures morethan simply "maturation" for it incorporates manypersonality characteristics which earlier we have

2The OPI instrument contains 385 true-false items relating tostudents' opinions, attitudes, and feelings on a variety of subjects.Validation dais consist primarily of correlations with other person-ality measures including: Allport, Vernon, and Lindzey Study ofValues; Strong Vocational Interest Blank; Guilford-ZimmermanTemperament Survey; Opinion, Attitude, and Interest Survey;Myers-Briggs Tme Indicator; Kuder Preference Record; CaliforniaPsychological Inventory; Edwards Personal Preference Schedule:Activities Index; and the Minnesota Multiphaic Personality Inven-tory (Heist & Yonge, 1968).

Reliability figures Computed by the Kuder-Richardson Formula 21and the corrected split-half method estimate the internal consis-tency of various OPI scales ranging from .67 to .89. The majority ofthe test-retest reliability coefficients are above .85 (Heist & Yong*.1988).

In addition the OPI has had extensive validation through its use inthe study of effects of colleges on students. For example: Beech(1966; 1967); Elton (1969); Elton and Rose (1968); Flacks 09834Heist (1960);,Katz (1988); Korn (1967); McConnell et al. (in Press);Snyder (1967); Stewart (1984); Trent (1964, 1967); Trent andMedsker (1968); and Tyler (1963) havo mode use of the OPI in theirresearch.

1915

associated with competence in a learning society. Ahigh score on SM would indicate that an individualhas a tolerance for ambiguity; an open, receptivemind; flexibility; autonomy; and freedom fromauthoritarianism. An extensive discussion of theSocial Maturity scale of the OPI has been offeredby Trent and Medsker (1968, p. 278).

Scores from the OPI were available for thesample to be investigated in the present researchfrom testing in 1959 and retesting in 1963. Ameasure of change in social maturity could bedetermined by subtracting the SM score in 1959from the score obtained in 1963. Alternatives tothis technique have been offered by Lord (1956,1958) and Nichols (1967) and, in addition, Harris(1963) offers a compendium of thoughts on thematter. Much of these discussions is concernedwith improving the estimation of "true" changescores for individuals. However, since we are notprimarily interested in the "true" magnitude of thechange for particular individuals, but rather in therelative magnitude of change occurring withingroups of students under varying conditions ofsocialization, simple gain scores seem adequate forour purpose.

Precollegiate socialization background variables.Prior to entry into an institution of higher educa-tion the individual student's primary socializationagent was his family. Socialization experiences canbe seen as functions of the family's economic,political, and ecological structure; its place in thesociety's stratification system; and the presenceand salience of roie models (Inkeles, 1969).

The major early socialization agent outside ofthe family is, of course, the school. An importantfactor in school socialization is the importance ofmeasured ability in terms of the kinds of oppor-tunities and experiences with which the student isconfronted (Hodgkins & Herriott, 1970; Schwebel,1968). In addition, the conjunction .of socializationwithin the family, the community, and otherorganizations such as the church leads to alternativemodes of socialization within the school setting.

The most perfect representation of an individ-uars precollegiate socialization background wouldbe the individual's entire precollegiate socializationhistory. However, an adequate model for precol-legiate socialization background can be achieved if

.4 .

proxy variables are available for the experienceswhich occur within the major socializationagenciesthe family, the church, and the school.

The precollegiate socialization backgroundvariables which follow constitute a model forsocialization experiences prior to entrance intohigher education. Variables which are proxy forcommunity, family, and idiosyncratic factors arerepresented in the model. The following variablesare included:

B1 = S's sexB2 = S's father's occupationB3 = S's religionB4 = S's hometown population136 = S's SCAT score or equivalent measure of

cognitive abilityB6 = Student enrollment of S's high schoolB7 = S's high school class rank.

The student's sex is proxy for the intricatesocialization patterns that vary because of theascribed sex role. The expectations, the attitudesand abilities, and many other socialization experi-ences are differential in our society purely on asexual basis. The student's family socialization canperhaps be best represented by determining thefamily's position in *he stratification system.Father's occupation has been shown to be aneffective proxy for a vast array of phenomenawhich differ depending upon the place of anindividual's family in the stratification system.Ranging from the purely physical such as theadequacy of health care to the inculcation ofpolitical attitudes, the effects of stratification uponthe individual are great (Caplow, 1964; Kahl,1957). In addition, religious affiliation can be seenas proxy for variations in world view that affectthe socialization goals and outcomes of the family(Rhodes & Nam, 1970).

The community in which the student is raisedhas particular salience for the type of socializationexperiences that the student undergoes. The partic-ular agencies and resources that are available andutilized during the student's precollegiate sociali-zation experiences are quite divergent dependingupon the community. One indicator of theenvironment of the community in which thestudent is socialized is its population. Theresources, the potential experiences, the social

16 20

agencies, the pace of life and expectations regard-ing behavior will vary depending upon the popula-tion of the student's hometown (Duncan & Reiss,1956; Haller & Wolff, 1965; Schnore, 1966).

The remaining three background variables areproxy for school socialization. Measures of cogni-tive ability determine to a large extent not oniy thesocialization goals of the school but the ability ofthe student to accept varying socialization pro-cesses. The findings of Barker and Gump (1964)provide important evidence that high school enroll-ment has important consequences fur the behaviorand experiences of students. High school size canbe seen as proxy for such factors as the variety ofbehavioral settings in the school, the degree of

differentiation in the curriculum and the scope ofextracurricular behavior.

Finally, the degree to which the individualstudent has achieved the objectives of a cognitivenature which the school socialization experiencesare designed to promote can be represented byclass rank. Since the student is rewarded primarilythrough grading for meeting socialization demands,consistent high grades, which would result in ahigher class rank, would represent a consistency inmeeting socialization expectations of the school.

In summary, Table 5 shows how precollegiatesocialization background variables within theTrent-Medsker data can be viewed as proxies forparticular elements within the accounting scheme

TABLE 5

Elements within the Accounting Scheme for the Structural Analysisof Precollegiate Socialization Background Matched with Background

Variables Which Can Be Viewed as Reflective of Thosrt Elements

Socialinstitution

F AM I LY

Structure

Authority StructureKinship StructureValue Structure

Knowledwi StructureEconom!c StructureEcological StrUcture

Affective StructureReward & Sanction System

SCHOOL Authority SystemReward & Sanction SystemGoal StructurePersonnel SystemEnvirönmental ExchangeEcological System

Interaction System

RELIGION Belief'SystemValue SystemDifferentiation of Activity

e,.,f7

17 21

Variable

B1 S's Sex

B1 S's SexB2 S's Father's Occupation

B2 S's Father's OccupationB2 S's Father's OccupationB2 S's Father's OccupationB4 S's Hometown PopulationB

1Qex

B1 S's Sex

B7 S's High School Class RankB7 S's High School Class RankB6 Student Enrollment

,

B4 S's Hometown PopulationB4 S's Hometown PopulationB6 Student EnrollmentB6 Student Enrollment

83 S's Religion83 S's Religion

presented for the structural analysis of precol-legiate sncialization background (Table 3). Forexample, B2, father's occupation, can be seen asproxy for the economic, ecological, knowledge,and value structure of the family because it can beused as a measure of the socioeconomic position ofthe family.

Operationalization of the above variables is asfollows. The respondent's sex, father's occupation,and religion were determined by the followingitems from the "High School Senior Question-naire" which was administered at the participatinghigh schools in the spring of 1959.

B1 = S's sex: Sex 1 Male 2 Female

B2 = S's father's occupation: What is your father's occupation?What does he do? Be as specific as you can. TELL US EXACTLYWHAT HE DOES AND THE KIND OF PLACE WHERE HEWORKS. For example, "sells clothes in a department store.""Operates lathe in a machine shop." "Is an office manager in aninsurance office." (lf he is dead, say what his occupation was.)

B3 = S's religion: What is your religion? (Answer only if you wish.)Protestant (DenominationRoman CatholicJewishOther (Write inNone

High school class rank (B7) and SCAT scores(Be) were obtained from official school records.Aptitude scores were available from all the 37participating high schools, although the scoresrepresented 18 forms of 11 different tests admin-istered at various times between the ninth andtwelfth grades. The method of equating test scoresin order to arrive at a common measure whichwould indicate the relative aptitudes of the respon-dents may be found in Trent and Medsker (1968,p. 276). Since the range of SCAT scores fallingwithin each decile as a result of this procedurecompared favorably with the range of scores on thenational norms as reported in the manual, thederived scores are seen as a reasonableapproximation.

The enrollment of S's high school (B6) wasobtained by the "High School Information Sheet"which was completed by the high school principalin the summer of 1959. The item was,

18

How many students were enrolled in the tenth, eleventh, andtwelfth grades (combined) in the school year of 1958-59?

50-99100-249250-499500-749750-9991,000-1,4991 .500-1,9992,000 or more

Finally, S's hometown population (54) wasobtained from records at the Center for Researchand Development in Higher Education. This infor-mation was obtained originally from RandMcNally's 1958 Road and Reference Atlas.

College structure variables. This researchproposes to make a preliminary test of theadequacy of the accounting scheme for the studyof the structure of institutions of higher educationwhich was presented in Table 2. The ideal test ofthe scheme would be to obtain measures ofvariables which would be proxy for each of thesystems which can be used to differentiate institu-tions of higher education. Upon the fullexplication of the various systems, multivariateanalyses could be carried out to determine thepredictive powers of the scheme with respect tothe production of competence in. a learningsociety.

The present test of the scheme will be con-ducted with a more modest model utilizing vari-ables which were accessible from the Trent-Medsker data or obtainable through standardreference volumes. Each of the proposed collegevariables is to some degree a proxy for character-istics which would be descriptive of the varioussystems in the accounting scheme for the study ofthe structure of institutions of higher education.The college variables which will be utilized for thispreliminary test of the scheme are:C1 = Academic quality of the college or universityC2 = Undergraduate enrollment of the college or

un iversityC3 = Type of organization of the college or

universityC4 = Type of control of the college or universityC5 = Population of the college communityC6 = Intensity of collegiate socialization setting.

These variables represent an initial attempt toexplicate indicators for the proposed accounting

22

scheme for the study of the structure of institu-tions of higher education. Academic quality (C1)can be seen as proxy for characteristics of four ofthe systems that compose the schemethe rewardsystem, the personnel system, the ecologicalsystem, and the environmental exchange system.

With increased academic quality comes increasedprestige. This prestige represents reward in terms ofthe recognition and social deference given to thestudents and faculty of institutions of high quality.In addition, monetary rewards accompany prestigein terms of increased visibility of faculty and theselection process which is utilized by prospectiveemployers of college graduates. Quality reflectscharacteristics of the personnel system since thedegree of professional training and the relativeavailability of faculty personnel to studentpersonnel are components of quality. In addition,the quality of student personnel that the college oruniversity attracts will vary with academic quality.

The ecological system is reflected in the measureof academic quality in physical plant and amountof funding per student. Finally, the environmentalexchange system is represented via the fact thatcolleges and universities are very aware of theirplace in the stratification system of higher educa-tion, and quality will be a reflection of the locationof any given institution within that stratificationsystem. In addition, any number of benefits willaccrue to an institution of higher academic qualitythrough interchange with alumni, business, andgovernment.

The undergraduate enrollment (C2) can be seenas proxy for three of the systems within theproposed scheme. The most obvious connection, ofcourse, is between enrollment and the personnelsystem. In addition to being purely descriptive ofthe personnel system, enrollment will also beproxy for the differences in the types of relation-ships and interaction between faculty and student,the general climate of interaction and phenomenasuch as previously described with respect to high

school size. Enrollment is also proxy for featuresof the ecological system. It is reflective of suchfeatures as the number of buildings, the density ofpersonnel, difficulty of obtaining isolation, and anumber of more subtle ecological features.

Type of college or university (C3) is particularlysalient for the goal system but certain features ofthe interaction system, the personnel system, and

I

19

the environmental exchange system can also berepresented by the type of institution. The diver-sity of goals can be particularly distinguished whendifferentiating between a university and a college.In addition, if type of college also includes thefurther distinction between denominational andnondenominational institutions, the nature of thegoal structure is further revealed. The most readyexample of how type of institution can be proxyfor the interaction system is the fact that majoruniversities have a different personnel structuresince, in addition to undergraduates and faculty,there are graduate students. This addition ofpersonnel is reflective of differences in allocatingenergy in terms of undergraduate teaching, forexample.

The personnel system is further reflected in typeof institution via the fact that major universitiestraditionally attract a more professionally trainedfaculty and also possess personnel for researchwhich colleges do not. Furthermore, for denomina-tional institutions the degree of religiosity may besomewhat reflected in type of institution. This isillustrated by the small denominational collegestaffed by clergy.

Type of institution can also be reflective of theenvironmental exchange system. A particularexample would be the kinds of environmentalexchanges between state colleges or public univer-sities and government agencies. This exchangewould not only be in the realm of policy but alsofinance and exchanges with specific governmentalagencies for other types of services.

Type of control (C4) is closely related to type ofinstitution in terms of being proxy for character-istics of systems within the accounting scheme.The authority system is reflected in terms of thegoverning bodies and types of authority structuresthat are present with variations in control. Publicinstitutions usually have a control board that ispolitical in nature whereas nondenominationalprivate schools are more closely representative ofalumni. Religiously controlled schools, of course,have an increased emphasis on religious and clericalmatters and personnel. Reflection of the environ-mental exchange system somewhat overlaps thepreceding discussion because with the variations insupport and control come variations in the natureand source of exchange between the institutionand its environment. For example, the relationship

23

of the institution to governmental agencies wouldvary greatly depending upon the nature of controlof the college or university.

Population of the college community (CO isparticularly related to the ecological system in thatthe size of the community in which the institutionis located is reflective of a particular aspect of theecological setting of the college or university. It isreasonable to expect the relationship between thecollege or university and its surroundings will varywith population. Institutions in a city of over amillion people will have different relationshipswith their surroundings than do institutions intowns of 10,000. In addition to the relationship ofthe institution to its surroundings, the potential forvarying relationships between the personnel of thecollege or university and the surrounding commu-nity differs with increased population.

Finally, the remaining college variable, intensity(C6), is particularly reflective of. the interactionsystem. Intensity refers to the extent to whichthere are competing social structures that lieoutside of the socialization environment of thecollege or university to which the student isexposed. Intensity would be extremely high, forexample, within a social setting that could beconsidered a "total" institution (Goffman, 1959);whereas in a situation where an individual was onlyhaphazardly exposed to a socialization agent, theintensity of exposure would be low.

In summary, Table 6 shows how the collegevariables available within the Trent-Medsker dataand other reference sources can be viewed asproxies for particular systems within the account-ing scheme offered in Table 2 for the study of thestructure of institutions of higher education. Forexample, the authority system is reflected incollege variables 3 and 4, type of institution andtype of control of the institution. The personnelsystem is reflected by C1, academic quality; C2,undergraduate enrollment; C3, type of institution;and C6, intensity.

To operationalize the college variables for thispreliminary test of the accounting scheme for thestructure of institutions of higher education thefol!owing procedures were used.

Although ideally one would like to measure thequality of a college by its product, the graduate, atthe current stage of development of data collectionin this area it would be a monumental undertaking

20

TABLE 6

Systems within the Accounting Scheme for the Studyof the Structure of Institutions of Higher

Education Matched with College StructureVariables That Can Be Viewed as Reflective

of Those Systems

Systems College variables

Authority System

Reward SystemSanction SystemGoal SystemInteraction System

Personnel System

Environmental Ex-change System

Ecological System

C3 TypeC4 ControlC

1Quality

C3 TypeC2 EnrollmentC3 TypeC

1Quality

C2 EnrollmentC3 TypeC6 Intensity

C1Quality

C3 TypeC4 ControlC6 IntensityC

1Quality

C2 EnrollmentC5 Col lege Community

Population

to obtain a measure which would reliably give usthis information for 683 institutions. For the timebeing, a more traditional notion of the quality ofeducational institutions will have to be used inorder to arrive at a measure of the stratificationsystem of institutions of higher education. Forpurposes of the present research an index, whichenables an investigator to go to any adequatelibrary and objectively rate the academic quality ofalmost any undergraduate college or university atvarious points in time, will be used.

Naturally some sacrifices and compromises haveto be made in building an index solely frominformation currently available in a library; but itis felt that the ultimate generality of such an indexwill be an advantage that will far outweigh anydisadvantage incurred from the loss of sensitivityto idiosyncratic features of an institution. This isespecially true in research designs such as the

2 4

present one where there is a need to consider asmany as 683 colleges.

A thorough search of the literature in this arearesulted in the identification of numerous indica-tors of academic quality that could be incorpo-rated into such an index (Pavalko & Ullrich, 1970).Six of these indicators were eventually selected forthe index. These six seemed to be supported byconsiderable consensus in the literature with regardto their validity. They also utilize informationwhich can be obtained from two standard refer-ence volumesCass and Birnbaum (1966) andCartter (1964). The six indicators include:

Faculty salariesSelectivityNumber of volumes in the libraryPercentage of doctorates on facultyRatio of library books to studentsRatio of students to faculty.

The rationale for inclusion and exclusion ofvarious indicators follows.

Faculty salaries are of obvious importance in thecapacity of a college or university to attract highcaliber academic personnel. It seems reasonable toassume that colleges and universities which payhigh faculty salaries will have, in general, highercaliber academic personnel than those that pay lowsalaries. Therefore, these institutions tend to be ofhigher academic quality as a result of the highercaliber of faculty.

Selectivity was included in the index from ameasure developed by Cass and Birnbaum and isbased upon information such as the percentage ofapplicants accepted by the college, the average testscores of the recent freshman classes, and otherrelated data which measure the scholastic potentialof the student body. As Cass and Birnbaum (1966)put it,

this index is a crucial measure of the academic quality of a collegebecause, as current research on higher education indicates, eninstitution of hillier learning can never be much better than itsstudent bodyand is not likely to be much worse (p. xvi) .

Both the size of the library and the ratio oflibrary books to the number of student's haveattained wide acceptance as indicators of academicquality. Lazarsfeld and Thielens (1958) acknowl-

.C,

21

edged this when they stated,

*le e +.1

the absolute size of the library is a valid indicator of quality sincelibrary books are in principle available to all students and the sheernumber of books suggests the breadth of scholarship possible. Atthe same time, the larger the ratio of books to students, the moreaccessible are library materials to each individual (p.411) J.

Proportion of doctorates on the faculty andstudent/faculty ratio are also widely acceptedindicators of academic quality. Lazarsfeld andThielens used proportion of PhDs and Jencks andRiesman (1968) call attention to the validity ofstudent/faculty ratio as an indicator of institu-tional quality. Information concerning both ofthese indicators is readily accessible; therefore,both indicators are included in the index.

Aptitude level is included in the index since Cassand Birnbaum used it as one of the selectivitycriteria. It is unfortunate that per student expendi-ture could not be included in the index because itis still another widely accepted indicator ofacademic quality. However, this indicator did notmeet the accessibility requirement of the index andis therefore excluded. The size of the student bodywas discarded as an indicator of quality because oflack of supporting evidence for its validity. Finally,the number of course offerings was discarded as anindicator because it did not fulfill the accessibilityrequirement.

The quality scale was operationalized using thefollowing procedures. Faculty salaries were scoredutilizing the AAUP salary rating (AAUP, 1964) forthe average salary rating of a given institution.Ratings of A and B were assigned a value of 4, C avalue of 3, D a value of 2, and E, F, and G a valueof 1. For those institutions which were not rated,Appendix Table C was used in order to assign themodal rating for institutions of a given type, bytype of control and geographic region.

Selectivity of a college was determined using theselectivity rating of Cass and Bimbaum (1964).The most selective schools were assigned a value of4 and the least selective schools a value of 1, withintermediate values of 3 and 2. If the selectivityrating was unavailable, the percentage of theapplicants accepted by the institution was utilizedas a substitute measure of selectivity. Assumingthat Cass and Birnbaum had rated the veryselective schools due to their prestige, it followsthat those schools not rated would be of a lower

selectivity. Therefore, if a school accepted morethan 50% of its applicants, it was rated as 1, and ifthe school admitted less than 50% of the appli-cants, it was rated as 2. Percentage of applicantsaccepted was obtained either from Cass and Birn-baum (1964), Canter (1964), or The College BlueBook (1965).

The size of the library was determined by datafrom Cass and Bimbaum (1964), Cartter (1964), orThe College Blue Book (1965), or the collegecatalog. For the "quality scale" the number ofvolumes in the library was divided into quartersand assigned values as follows:

4 = 321,000-5,000,000 volumes3 = 121,000-320,0002 = 61,000-120,0001 = 60,000 or less.

Total number of full-time u ndergraduates,

number of full-time faculty members, and numberof faculty members with doctorates were obtainedfrom the same sources. For purposes of con-structing the scale percentage of faculty holdingdoctorates, ratio of library books per full-timeundergraduate student, and ratio of full-time stu-dents to number of faculty members were dividedinto quarters and assigned the following values:

Percent of faculty with doctorate:4 = 50-99%3 = 36-492 = 25-351 = 7-24

Books per student:4 = 111-999 books/student3 = 67-1102 = 46-661 = 5-45

Students per faculty member:4 = 1-9 students/faculty member3 = 10-122 13-151 = 16-28.

The resultant scale of quality was attained bysumming the six scale values and ranged from ahigh quality rating of 24 to low rating of 6.

22

Population of the college community was deter-mined either from the population for the commu-nity in which the institutiol was located asreported by Cass and Birnbaum (1964), or fromthe 1960 Census.

The Research Sample

For this preliminary test the population wasdefined as those students who entered a regular4-year college or university on a full-time basisupon graduation from high school and whoattended only one such institution over a 4-yearperiod.

The original Trent-Medsker longitudinal sampleavailable for this research consisted of 2,404students who had entered higher education fulltime in the fall of 1959. Of the original 2,404'students, 681 had gone to two institutions ofhigher education, 107 had gone to three, and 2 hadgone to four schools. Of the remaining 1,614students who had attended only one institution ofhigher education, 362 attended a school which wasnot a regular 4-year college or university (forexample, many attended junior colleges). There-fore, the sample available for analysis consisted of1,252 students who had entered a regular 4-yearcollege or university on a full-time basis directlyupon graduation from high school and whoattended only one such institution over a 4-yearperiod. Missing data necessitated deleting 56 casesfrom the sample so that the actual analyses werecarried out on a sample of 1,196 students.

Design for Analyses

Two variables, in addition to those noted above,are central to both the theoretical framework ofthis research and the design for analyses of thedata. They are duration of exposure to collegiatesocialization (D) and original wore on the SM scale(SM0). In the ideal analysis, the data would bepartitioned on each of these variables and cross-classified to create first- and second-order sub-samples. The data would be partitioned on dura-tion of exposure because of the theoretical impor-tance of duration of exposure to socializationagencies and changes in personality. Partitioning onSM0 is essentially an empirical step necessary dueto the ceiling effect of the SM scale. The ceiling

26

effect of any scale prevents an individual whooriginally scored high to show as much gain as anindividual who had originally scored low. Inaddition, it can be argued that a given gain at thehigh end of a scale such as SM could have adifferent meaning conceptually than the samedegree of gain at the low end of the scale.

Regression analyses involving large numbers ofindependent variables for small samples have beenshown to be unstable in nature (Coleman, 1964).Since formal statistical techniques will be used totest the two hypotheses, the type of partitioningdescribed previously would likely result in smallsample size for some of the subsamples and thusunstable estimates of population statistics; hence,the total sample will be used. However, a pro-cedure for overcoming these problems will beintroduced in a later section to elaborate theformal tests of hypotheses and to enhance ourunderstanding of the data.

Hypothesis One stated that student personalitychanges are due to the social structures of thecolleges the students attend. In order to testHypothesis One, a regression equation which repre-sents the change in social maturity as a function ofthe college variables will be developed.3 Upon thecalculation of the regression equation, the multiplecorrelation coefficient (R) will be determinedbetween the dependent variable and the "best"weighted sum of the independent variables.

An analysis of variance will then be conductedto determine the statistical significance of themultiple correlation coefficient. Support for thishypothesis will be claimed if the R is statisticallysignificant at the .05 level.

In order to carry out the regression analysis it isnecessary to create dummy (i.e., binary) variablesfor those variables below the interval level. Dummyvariables were created where necessary followingthe method described by Draper and Smith (1966).For example, for C3, type of college, a dummyvariable was created for "state college" by assign-ing to each student a value of "1" if he attended astate college ad a value of "0" if he did not attenda state college. In each case where dummy variableswere necessary, one category of the original vari-able was designated the "0 cell" and not enteredinto the regression equation.

Next we desire to test the assumption thatprecollege socialization is related to student

23

rzolevomrennr*,-,T1fr'

personality change in college. We plan to test thisby constructing a regression equation which repre-sents the change in social maturity as a function ofthe precollegiate socialization background vari-ables. Upon calculation of this regression equation,the R for the 'test" weighted sum of theindependent variables will be calculated.' An

3 The regression equation for testing the first hypothesis is,

SM = a + be + cC + dee ecdc fc ndc gcpu + hCPdu9 1 2 .11 3 3 3 3

u,PUb 6,w ,Where: SM = change in social maturity scale

a, b, c, . m = regression coefficients

CI = quality

C2 = enrollment

Cne = dummy for state college3dc

C3 = dummy for denominational college

Cnde = dummy for nondenominational college3

CPu = dummy for public university3

CPdu = dummy for private denominational university3tcndu,j = "0 cell"nondenominational private3

universities (not in equation)

,pub = dummy for public controls'4epro = dummy for Protestant control4Cent = dummy for Catholic control4

[C11 = "0 cell"private control (not in equation)

C5 = population of college community

C6 = intensity of collegiate socialization setting

4The statistical model utilizing only precollegiate socializationbackground variables can be represented by the regression equation:

SM = a + bB + cB + dl3Pte + + Brew + + h8 + 1B1 2 3 3 3 4 5 6

+ j87 .

Where: SM9

a, b, c, . . . j

B.

= change in social maturity scale

= regression coefficients

= femaleness

B2 = socioeconomic status as measured by

father's occupation

27

= dummy for Protestant religious affiliation

.= dummy for Catholic religious affiliation

[continued]

*".1..1.,a,71.!...Mt11,29Strnr.i5rnrS_

analysis of variance will be carried out to ascertainthe statistical significance of the R and if it is

significant at the .05 level of probability, supportfor this assumption will be claimed.

Hypothesis Two holds that student personalitychange is a function of the college social structureafter controlling for precollege socialization. Thetest of Hypothesis Two requires a slightly morecomplex statistical procedure. We shall computethe multiple correlation coefficient for the collegestructure and precollegiate background variablestogether and compare it to the multiple correlationcoefficient for only the precollegiate backgroundvariables. In this way, the total predictive ability ofan analytical model utilizing only precollegiatesocialization background variables can be com-pared to that of a model utilizing college structurevariables in addition to precollegiate socializationvariables.

To make this comparison, the multiple correla-tion coefficient produced from the regressionequation involving only precollegiate socializationbackground variables will be compared to amultiple correlation coefficient produced from aregression equation involving college structurevariables as well as the precollegiate variables.' TheF test will be utilized to determine the statisticalsignificance of the difference betweer; these twomultiple correlation coefficients (Baggaley, 1964).Support for the hypothesis will be claimed if the Rfor the background and college variables is signifi-cantly greater statistically at the .05 level than theR for the background variables only.

However, as noted by Feldman (1970), Nichols(1967), Stanley (1967), and Werts and Watley(1968), a statistical model of this type tends tounderestimate the effects of the collegiate experi-ence, for it assigns to the precollegiate variables all

of the variance in personality change which theyshare with collegiate variables. Unfortunately,statistical models which can unambiguously sepa-rate the effects of precollegiate and collegiateexperiences within a modified longitudinal designof -the type used by Trent and Medsker incollecting their original data are not currentlyavailable. However, since the statistical modelwhich we will use is clearly a conservative one,given our theoretical interests, we are far morelikely to accept the null hypothesis associated withHypothesis Two in error than we are to reject it in

24

+........

error. Thus, we can have considerable assurancethat any support which we can claim for Hypothe-sis Two using this model is extremely likely to besubstantiated by any subsequent research with alongitudinal design covering the period from birthto completion of college.

Results

Formal Test of the Hypotheses

Hypothesis One. A regression analysis wasconducted on SMa for the total sample utilizingthe college structure variables as independentvariables. Support for Hypothesis One is shown bythis regression analysis. A multiple correlationcoefficient (Re) of .1614, which is statisticallysignificant at the .01 level, suggests that the degreeof personality change can be seen as a function ofthe college structure variables.

Assumption. The regression analysis involvingonly precollegiate socialization background vari-ables produced an RB of .1861, which is statisti-

Blew

foothi1°3

B4

86

137

w dummy for Jewish religious affiliation

w "0 cell"other and no religious

affiliation, (not in equation)

w population of S's hometown

SCAT or equivalent measure of cognitive

ability

student enrollment of S's high school

S's high school class rank

The regression equation which will be utilized to determine the

predictive power of the model containing both precollegiate

socialization background variables and college structure variablescan be represented by:

SM a + bB + cB + dBPr° + + fBiew + + + iB1 2 3 3 3 4 5 6

+ jB2 + kCi + 1C2 + mCic + nC(31c + oC + pCr + qCgdu

rcp scro rtt; + uCs + vC0

28

cally significant at the .001 level. Therefore,support for the assumption is claimed.

Hypothesis Two. A regression equation for boththe college structure variables and precollegiatesocialization background variables was constructedand this regression equation for the total sampleproduced an RB+C of .2412. This multiple correla-tion coefficient is significant at the .01 level.

To test Hypothesis Two, the difference be-tween RB+C and RB was calculated and thestatistical significance of the difference computed.For the total sample, the difference between thetwo coefficients is equal to .0551,6 thus support-ing Hypothesis Two. Clearly, we can increase thetotal explained variance in student personalitychange by utilizing an analytic model that containscollege structure variables as well as precollegiatesocialization background variables.

Discussion and Evaluation of the Results

It is useful to make a distinction betweenstatistical significance and theoretical significance,and then reinterpret our findings in light of theirsignificance for the theoretical framework whichhas guided this research. Given our objective, itseems appropriate to go beyond the formal tests ofsignificance to determine what kinds of addedinformation we can glean from the data.

As noted in the design of the study, duration ofexposure to collegiate socialization (D) and originalSM score (SM0) are important variables in ourresearch design, for each can possibly suppress (i.e.,attenuate) the true relationships between bothprecollegiate socialization and college structurewith personality change. In order to control forsome of the possible suppressing effects of thesevariables upon the relationships of primary inter-est, the hypotheses and assumptions will be recon-sidered within first- and second-order subsamplesthat can be created by partitioning and cross-classifying the research sample on these twovariables simultaneously.

Two first-order subsamples in terms of SM0 havebeen created by dichotomizing this variable at themedian. Five first-order duration subsamples werecreated by categorizing the number of years astudent was exposed to college, i.e., the number ofyears of enrollment before termination of hiscollege career. In this latter case, the first sub-

,y7.7: ^

sample is those students who were enrolled 1 yearor less. Duration of exposure for the secondsubsample is greater than 1 year but less than orequal to 2 years. Duration of greater than 2 yearsbut less than or equal to 3 years composes thethird subsample.

The fourth- and fifth-duration subsamples arethose students who stayed in college for more than3 years but did not receive a bachelor's degreeby the end of the 4th year and those students ofsimilar duration who did receive a bachelor'sdegree. It is assumed that for those students whodid stay in college longer than 3 years a meaningfuldistinction can be made as to the degree ofexposure utilizing the granting or not granting of adegree. It follows that, on the average, if anindividual student was fully exposed to the social-ization program, he would be certified as havingsatisfied the requirements of the socializationexperiences. However, in general an individual whowas not as completely exposed to the socializationexperiences for those years would not be certifiedas having met the requirements of the socializationagent. Therefore, the granting of a degree can beproxy for a further refinement in measuring thedegree of exposure to the socialization agent.

Ten second-order subsamples were obtained bycross-classifying the two categories of SM0 and thefive categories of D. Table 7 portrays the results ofthis refinement to our research design, summar-izing the total sample, the 7 first-order subsamples,the 10 second-order subsamples, and their associ-ated number of data cases.

Because of the instability of the regressionanalyses carried out with large numbers of indepen-

6To calculate the statistical significance of this difference, thefollowing formula is used (Bewley, 1964, p. 200):

F =

032+CR2

'(SS No. of B variables No. of C variables 1)

B B

(No. of B+C variables No. of 13 variables) R

Computation for this F test is

(.058 .035) (1196 10 24 1) 26.703F = 1.92.

(24 10) .058 13.942

Referring to the typical "F-table," we find that an F as large as 1.92with the given degrees of freedom has a chance probability of lessthan .05. SS = Sample Size.

25 29

7,-veancsranvorrprmremmmasPryx,r, sekl,,,.

dent variables on such small samples as are re-ported in Table 7, a procedure for estimating the"true" multiple correlation coefficient for smallsamples was desirable. The typical procedures forcorrecting R for degrees of freedom seemedinadequate; its estimate is based solely upon theparticular subsample in which it is computed.Therefore, we chow a procedure for poolingdegrees of freedom similar to that used by Se !yin(1960) to assess the consistency of results in astudy of the effects of leadership.

TABLE 7

Frequency Distribution of First- and Second-OrderSubsamples Formed by Categorizing and Cross-Classifying Original Social Maturity Score (SM0)

and Duration of Exposure to College Structure (D)

Duration (D) ofexposure to collegiatesocialization, in Original SM Total

years Low High

D < 1

1<D<.2 66a

2<D<.3 41a

3<D<.4 152a

no degree3<D<4 280a 288a 568b

degree

0 < D < 4 634b 562b 1196c

95a 61a63a29a

121a

156b

129b

70b

2731)

a = second-order subsampleb = first-order subsamplec = total sample

To arrive at an estimate of the "true" multiplecorrelation coefficient within subsamples of I ourdata, "pooled" multiple correlation coPfficientswere determined. The pooled Rs were calculatedby computing the average of the multiple corre-lation coefficients within either the first- orsecond-order subsamples. To arrive at the pooled Rfor first-order subsamples, the arithmetic averageof the Rs for each category of D was calculatedwithin each category of SM0.

To arrive at an estimate of the "true" multiplecorrelation coefficient for a typical second-order

26

subsample, where theoretically we expect toexplain the greatest degree of variation, a "grandaverage" of the Rs fo. the 10 second-ordersubsamples was computed. This "grand average"will be utilized to estimate the significance of ourtheoretical framework.

Hypothesis One. Table 8 presents the results ofthe subsample analyses for the relationship of thecollege structure variables with personality change.There it can be seen that whereas the R for thetotal sample is .16, the Rs for the first-ordersubsamp les range from .13 to .44. Similarly, the Rsfor the second-order subsamples range from .14 to.63.

Table 9 compares, for the total sample and therelevant subsamples, the originai Rs with thoseobtained from the pooling process described previ-ously. Of particular interest is the "grand average"

TABLE 8

Multiple Correlation Coefficients forCollege Structure Variables on Change

in Social Maturity for First- and Second-Order Subsemples and the Total Sample

Duration (D) ofexposure to collegiatesocialization, in Original SMyears Low High

Total

D 1 year .40 .49 .30(95) (61) (156)

1 < D 5 2 years .63 .34 .42(66) (63) (129)

2<D<3 years .50 .58 .44(41) (29) (70)

3 < D 5 4 no degree .14 .29 .13(152) (121) (273)

3 < D < 4 degree .22 .28 .20(280) (288) (568)

< D <4 years .19 .25 .16

(634) (562) (1196)

Note: Number in parentheses = sample or subsample sizes.

30

TABLE 9

Pooled Multiple Correlation CoefficientsCompared to First-Order Subsample and Total Sample

Multiple Correlation Coefficients for C1.6 on SM9

Subsampledesignation Original Rs Pooled Rs

Low Original SM .19 .38High Original SM .25 .39D 1 .30 .44

1 < D 2 .42 .492<D-.<-3 .44 .543<D-<-4

no degree

.13 .21

3 < D -<.. 4

degree

.20 .25

First-Order SM .21

SubsamplesFirst-Order .30

DurationSubsamples

Total Sample .16All Second-Order .39

SubsamplesGrand Average

R computed across the 10 second-order sub-samples. This coefficient of .389 is considerablylarger than the R of .1614 based upon the totalsample and used for the formal statistical test ofHypothesis One. As expected, when controlling forSM0 and D, the explanatory power of an analyticalmodel containing only college structure variables isincreased. Thus, 14.9% (the square of .39) of thevariation of change in social maturity can beexplained by a model containing only collegestructure variables within a typical second-ordercell, assuming our estimate of the "true" multiplecorrelation coefficient is accurate.

These measures of six admittedly unrefinedindicators of social structure explain a significantamount of the total variation in change of socialmatuilty.

Assumption. By utilizing a procedure identicalto the one above, the assumption can be shown tobe theoretically significant to our research model.From multiple correlation coefficients in Tables Aand B of the appendix, the grand average for 81, 2,

7 on SMg can be seen io be .38, considerably. . .

J. 0,

."*."-w.errle,,,,rers0.1.,,Ant",". Verr*FI'Vr.7:11;7',Y.177.47'"

larger than the .1861 used to formally test theassumption. Squaring the grand average gives anestimate of the true amount of variance explainedin a typical second-order cell utilizing an analyticmodel containing only backgro6nd variables.Squaring .38 gives an estimate of 14.6% of thevariance which can be attributed to a modelcontaining only background variables.

Hypothesis Two. In order to contrast theexplanatory model containing both backgroundand college variables with the models containingonly background or college variables, multiplecorrelation coefficients from Table C in theappendix can be used to compute the grandaveraye of RB.I.0 for a typical second-order cell.Table D 4.1 the appendix shows that a grand averageof .51 is computed for the multiple correlationcoefficient involving both background and collegevariables. Squaring this estimate of the truemultiple correlation gives a figure of .258, whichcan be interpreted that 25.8% of the variation inchange in social maturity can be attributed to amodel containing both precollegiate socializationbackground variables and college structure vari-ables.

To estimate the "true" manner in which Vc andVb overlap, it is necessary to compute for eachsecond-order cell the difference between theamount of variation explained by a model contain-ing both background and college variables and (a)the amount of variation explained by a modelcontaining only background variables, and (b) theamount of variation explained by a model con-taining only college variables. Tables F and I in theappendix give the calculation of these differencesfor the first- and second-order subsamples and thetotal sample.

From these differences, estimates of the truedifference for a typical second-order subsample canbe established through the pooling process. TablesH and J of the appendix show this pooling process.

2 2The grand average for R B4.c REIN .113, and the2grand average for RB+C R6 is .117. The

difference of .113 represents Vb+c Vb or, in

other words, the amount of explained variation in

Vc which is not within the intersection of Vb andVc but is within the union of Vc and Vb. Likewise,the difference of .117 represents Vb+c Vc which

is the amount of explained variation in Vb which is

27

ai

not in the intersection of Vb and Vc but is withintheir union.

Figure 3 is a Venn-diagram representinc, ourestimate of the true overlap of Vb and V. Drawnto scale, Figure 3 shows that 11.3% of theexplained variation in social maturity is encom-passed by the part of Vc which is not overlappingwith Vb; on the other hand, 11.7% of theexplained variation in change in social maturity isencompassed by the part of Vb that does notoverlap with V. The intersection of Vb and Vccomprises 2.8% of the explained variation in

change in social maturity.This diagram shows that the theoretical model

which we have proposed has merit when studyingthe effects of higher education on student person-ality. The fact that Vc is not a subset of Vbdemonstrates that changes in personality occurringduring the college experience are not likely to bemerely changes that occur because of precollegiatesocialization. Indeed, the overlap of 2.8% suggests,as we have previously stated, that precollegiatebackground does affect the type of collegiatesetting to which students matriculate; but thatonce within the collegiate socialization setting,socialization outcomes of the college experience

Limitations

"mrdrrOVIISOMATIVerFIF95L.,

vary independently with the structure of theinstitutions of higher education.

Fig. 3. The amount of total variatior in personality changeattributable to a model containing both background andcollege variables so that their intersection is apparent.

Conclusions

Research concerned with the structure of insti-tutions of higher education and socialization tocompetence based on our theoretical frameworkwould have been free of limitations to the extentthat the ideal research design presented earlierunder Design of the Study had been followed. Thesteps in such an "ideal design" would include:

1. Randomly selected students from randomlyselected colleges and universities.

2. Students measured for competence at entranceto higher education.

3. Complete knowledge of the students! precol-

legiate socialization background.

4. The social structure of the institutions of highereducation would be completely explicated.

5. The intensity and salience of various elements ofthe structure of said institutions for individualstudents would be determined.

6. Upon termination of their collegiate career,students' competence would again be measured.

In order to make explicit the limitations of thisstudy, the design used here may be contrasted tothe ideal design proposed in Design of the Study.

Due to the fact that secondary analysis wasutilized for this preliminary test of the theoreticalframework, step one (random selection of studentsand institutions of higher education) of the idealdesign was not followed. The original purpose of

28 32

147""..117:7;77-71.TI7Menitnyfmn.M.,Trgrt,IrenIsien .

the Trent-Medsker sample was to study the rela-tionship between the type of higher educationinstitutions available in a city and entrance ofstudents in those cities to higher education.Because of the nonrandom selection of the stu-dents and institutions of higher education, thegeneralizability of these research findings is

limited. However, since the cities selected did varya great deal, especially with respect to geographiclocation, it seems reasonable to assume that thestudents studied in this preliminary test represent alarge section of students who do enter highereducation. Although no formal generalization tosuch a population is possible, there seems to belittle need to reject these findings because of thenonrandom sample.

Items 2 and 6 of the ideal research design areconcerned with the measurement of competence ina learning society as we have defined it. If one

n...aor,...AravermarmatProm.Vn...1,1r,ILT:Konv."-s,. r14.`1T9M"t.W4r.Mr4711.50'Wkier,'

allows the assumption that varying personalitypredisposes individuals to varying behavior, thechief potential limitation in measuring competencein the present study is the extent to which the SMscale correctly represents the accounting schemefor the study of personality.

By comparing the systems within the accountingscheme for the study of personality with person-ality attributes which the Social Maturity scalemeasures, an assessment of how completely SMrepresents the personality system can be made.Table 10 shows such a comparison. With theexception of "tolerance for ambiguity" whichmight be considered as representing temperamentin the "Psychomotor System," the SM scalerepresents only the cognitive mode of the "Modesof Functioning." This representation appears to bequite adequate for a preliminary study of the typeconducted here, but it is obvious that a full

TABLE 10

Systems within the Accounting Scheme for Personality Study Matchedwith Personality Traits Measured by the SM Scale That Can Be

Viewed as Reflective of Those Systems

Psychomotor System:

Idea System:

Motivational System:

Relational System:

Self-System:

TemperamentAptitudesSkills

I nformationOpinions and Attitudes

ValuesMotives and Needs

Authority FiguresIntimates and Peers

Collectivities

Conceptions of SelfModes of DefenseModes of Moral Functioning

Modes of Functioning: Cognitive Modes

Affective ModesConative Modes

+.` 2933

Tolerance for Ambiguity

Open, Receptive MindFlexibilityAutonomyNonauthoritarianism

representation of the personality system willrequire a consideration of many more facets of thestudent's personality.

At present, there does not appear to be either asingle psychological test, or a battery of such testswhich can give a comprehensive view of the manysubsystems of the personality system. Further-more, if such an instrument were available it wouldlikely report the personality system in somemanner other than a limited number of summaryscores. There is insufficient methodology foranalyzing a series of scores thus adding to thedifficulty of such a measure. It is suggested thatperhaps one way of overcoming these difficultieswould be to develop a scale to more directlyrepresent competence, perhaps in terms of sociallearning. Such a scale could provide a summaryscore that would be amenable to present tech-niques of data analysis.

In order to fully measure the students' precol-legiate socialization backgound, representation ofeach of the major systems to which the studentsare exposed (i.e., the familial system, the educa-tional system, and the religious system) would haveto be more fully explicated than was done in thispreliminary test. Although the background vari-ables which were utilized did represent each of thethree systems to some degree, there are obviousweak points in the representation of each of thesystems.

Along similar lines, Table 6 showed the mannerin which the college variables used in this researchrepresent various systems within the accountingscheme for the structure of institutions of highereducation. Even though each of the systems, withthe exception of the sanction system, is repre-sented by at least one proxy college variable, it isapparent that the indicators of college structurewhich were utilized in this research are very globalin nature. Such global indicators as the percentageof doctorates on the faculty or undergraduateenrollment are readily accessible and thus are verytempting. However, to gain a more accurate anddescriptive picture of the structure of institutionsof higher education, more refined measures ofstructure will be needed.

Some examples of the type of variables thatmight be considered as more refined indicators ofthe structure of institutions of higher education arefound in Table 2. For example, representing the

30

Authority System, such indices as the degree ofbureaucratization within the administrative frame-work of the college or university, the degree towhich decision making is on a collegial basis, andthe legitimacy which is given to authority figuresby various segments of the university populationshould be considered. Measures of such variableswould offer a much greater refinement than dothose indicators which were used in the presentresearch. Judging from the fact that our theoreticalframework was supported, using what are obvious-ly unrefined measures of structure, any addedrefinement in the representation of the structure ofinstitutions of higher education should greatlyimprove the explanatory power of the theoreticalframework which we proposed.

It is interesting to speculate as to the relativestrengths of the analytical model representingprecollegiate socialization background and theanalytical model of college structure. If one con-siders widespread usage of an indicator as indica-tive of the validity and reliability of that indicator,it appears obvious that the indicators utilized inthe precollegiate model offer a better representa-tion of precollegiate background than do theindicators of college structure. If we have repre-sented precollegiate background more successfullythan college structure, it is reasonable to assumethat if college structure were more adequatelymeasured, i.e., if better proxies were used, theamount of variance explained by college structurewould likely increase. If this is the case, we couldexpect the total amount of variation explained bycollege structure to be much greater than thevariation explained by precollegiate background,instead of approximately the same, as we havefound in this research.

A further limitation of this research can bediscussed in terms of the degree to which theindividual student is located within the structuralmatrix of his particular college or university.Within any social institution as complex anddiversified as today's colleges and universities,there is "room" for various individuals within thesame institution to move within and betweendifferent structural elements of the institution. Forexample, a student deeply involved in studentgovernment is likely to develop relationships withthe authority system that differ from those of astudent who comes to the campus only for formal

34

1

classes. Likewise, student contact with the person-nel system would vary depending upon his gradelevel and major area of study. 'his being the case,individuals can be differentially exposed to varicuselements of the structure of the institution andvarious elements of the structure can have differ-ential salience for different individual students.

Our attempt to locate the individual studentwithin the structural matrix of the institutionutilized simply the duration of his exposure to thecollegiate setting. However, this represents only aninitial attempt to locate the individual within thestructural matrix of the socializing agency. Morerefined measures of exposure to various structuralelements within the college or university, perhapsby means of time and motion studies or studentlog books, is needed along with a mechanism fordetermining the salience to the individual of thisdifferential exposure.

An initial attempt was made in this study todetermine the salience of the collegiate socializa-tion setting. Whether or not a student attended acollege in his hometown is admittedly an unrefinedmeasure of such phenomena. Improvements alongthis dimension might be made by attempting todetermine the degree to which a student identifieshis "socializers" as a reference group and whetheror not he possesses other reference groups whichwould be supportive of or detract from the goals ofsocialization to competence.

With the development of more refined measuresto locate the individual within the matrix of hissocializing institution a more refined method ofdata analysis could be adopted. Viewiag suchphenomena as extent of exposure and intensity ofexposure as control variables which could modifythe relationship between structure and the develop-ment of competence, these and other importantcontrol variables (for example, in this research thestudent's initial level of competence) could beentered into the regression equations utilized todetermine the relationship between structure andcompetence. In this way, a more systematicapproach to the possible suppressing effects ofsuch control variables could be implemented.

We have given a description of a much morerefined research design than has been utilized inthis research. However, we do not feel that in apreliminary test of a theoretical framework thelimitations which have been enumerated are

:.12

crippling in nature. Of course, as with any pre-liminary test, the results can best be viewed as aguide for future research. Furthermore, it seemsreasonable to consider this preliminary test in lightof its strong points. The Trent-Medsker dataoffered many advantages because of their longi-tudinal nature and each of the elements within theideal proposed was considered and developed tosome degree.

Implications

The development of our theoretical frameworkand subsequent preliminary test of that frameworkhave implications for two important areas of con-cern. The first is the study of higher education andin particular the study of the impact of highereducation on students. The second is a morecentral issue to the discipline of sociology. That isthe applicability of the theoretical framework tothe study of other socializing agencies and theeffects of these agencies on their "clients." Speci-fically, these other agencies could include theschools, prisons, mental hospitals, and the military.

The review of the literature on the impact ofcollege on students which was provided in thesections titled "Socialization, Social Structure, andCompetence" and "Approaches to the Study ofCollege Student Interaction with the Environ-ment" left little room for doubt that students dogrow in a direction consistent with competence ina learning society. However, much less agreementexists as to whether or not the college experience,per se, contributes to this growth or whether theselection processes of entrance into institutions ofhigher education merely provide a "place" fordevelopment, which has already been started, toreach fruition.

By providing a more comprehensive theoreticalframework than is present in previous research, thisstudy permitted an interpretation of data andresults in a much more meaningful way than hasbeen possible in previous work.. Our study offersstrong support for the hypothesis of an indepen-dent contribution of the impact of colleges anduniversities on the development of competence,after extraneous effects of precollegiate socializa-tion background have been accounted for. Theimplications for this approach in the study of otherkinds of effects of college are quite vast. The

theoretical framework provides a guide for thestudy of the impact of higher education in manyrealms of student behavior. One example of topicalinterest at this point in time would be the study ofthe radical student. Much more could be learnedconcerning the significance of the social structureof institutions of higher education to the develop-ment of radical behavior by utilizing the theore-tical framework provided herein.

The theoretical model developed in this study isnot limited to studying the effects of highereducation on students. The framework is readilyapplicable, with minor alterations, to any socialinstitution which is concerned with changingpeople. "People-changing institutions" include theschools at all levels, prisons, mental hospitals, andthe part of the military where training is mostprominent.

Within each of these social institutions existgoals oriented toward the modification of individ-uals. This modification may be minor as in the caseof the military or very drastic as in the case ofmental hospitals and prisons.

The fact that people-changing institutions are soprevalent in our society leads to a natural curiosityon the part of the social scientist; but in addition,the fact that large numbers of people are eitheremployed by or pass through such institutions makesknowledge of their outcomes important moregenerally. Since many of the outcomes of people-

changing institutions are integrally related toparticular value positions, understanding theireffect takes on added significance.

Consider prisons as an example of a people-changing institution. The value position implicit inthe sanctioning of individuals for unlawful be-havior is central to the institution's existence. Itseems obvious that at least in theory penal institu-tions wish to rehabilitate rather than to merelyconfine. The theoretical model presented in thisinquiry could be altered so that the developmentof habilitative behavior would be the focus for trl..structural analysis of penal institutions. Pre-penalsocialization background could be determined inmuch the same manner as precollegiate backgroundwas determined in this study. Of course, because ofthe fact that prisoners would likely be older thancollege students, their background would be moreextensive. However, our theoretical model wouldbe readily applicable to the study of such matters.

In conclusion, the applicability of the theore-tical model developed in this research appears tohave wide utility for the study of all types ofpeople-changing institutions. The structural analy-sis of the impact of such social institutions hasimportance for value positions, policy matters,administrative decisions, and for the expansion ofthe understanding of such institutions by socialscientists.

'11f4P.M"Tritlatert

TABLE A

Appendix

TABLE B

Multiple Correlation Coefficients forPrecollegiste Socialization BackgroundVariables on Change in Social Maturityfor First- and Second-Order Subsamples

and the Total Sample

Pooled Multiple Correlation CoefficientsCompared to First-Order and Total Sample Multiple

Correlation Coefficients for 1314 on SMg

Duration (D) of Subsampleexposure to collegiate designation Original Rs Pooled Rs

socialization, in Original SM Subtotal/years Low High total Low Original SM 23 .34

High Original SM .20 .42

D < 1 year .30 .50 .30 D < 1 .30 .40

(95) (61) (156) 1 < D < 2 .33 .432<D<3 .48 .62

1<062 years .43 .43 .33 3<D<4 .15 .23

(66) (63) (129) no degree3<D<4 .22 24

2<0 G3 years .45 .79 .48 degree(41) (29) (70) First-Order SM .22

Subsamples3<D<4 nodegree .25 .21 .15 First-Order .30

(152) (121) (273) DurationSubsamples

3<D<4 degree .29 .19 22 Total Sample .19(280) (288) (568) All Second-Order .38

Subsamples-0< D <4 years .23 .20 .19 Grand Average

(634) (582) (1196)

Note: Numbers in parentheses sample or subsample sizes.

TABLE C TABLE D

Multiple Correlation Coefficients forPrecollegiate Socialization Background

Variables plus College Structure Variableson Change in Social Maturity for First and

Second-Order Sub:ample: and the Total Sample

Duration (D) ofexposure to collegiatesocialization, inyears

Original SMLow High

Subtotal/total

D < 1 year .45 .63 .41

(95) (61) (156)

1 < D < 2 years .70 .54 .49MIN (63) (129)

2<D43 years .56 .86 .56(41) (29) (70)

3< D< 4 nodegree .30 .37 .18(152) (121) (273)

3 < D < 4 degree .35 .34 .30(280) (288) (568)

0<D44 years .28 .30 .24(634) (562) (1196)

Note: Numbers in parentheses - sample or subsample sizes.

Pooled RB+c Compared to Original First-OrderSubsamples and Total Sample RB+c

Subsample

designation Original Rs Pooled Rs

Low Original SM .28 .47High Original SM .30 .55D <1 .41 .521 < D <2 .49 .552<D43 .66 .703<D44

no degree

.18 .33

3<D44degree

.30 .34

First-Order SM - .29Subsamples

First-Order - .39DurationSubsamples

Total Sample .24 -All Second-Order .51

Subsamples-Grand Average

.lr'''T'?.YAI*14.5.ZIrtkVier3311=9;96ext-verissaakmonamramvarso. lavom.....ftemenmaniatalleaninselltIVIINKOMPAMMWrtr2;

TABLE E

Differences Between the Multiple CorrelationCoefficients for Both Background and College

Variables on Change in Social Maturity andthe Multiple Correlation Coefficients for Only

TABLE G

Pooled Differences Between RB+C and RBCompared to First-Order Subsemples and

Total Sample Differences

Background Variables on Change in Social Subsample Original Pooled

Maturity for First- and Second-Order designation differences differences

Subsamp les and the Total Sample Low Original SM .04 .13

High Original SM .10Duration (D) of D 4 1 .10 .14exposure to collegiate 1<D< 2 .16 .19socialization, in Original SM Subtotal/ 2<D<3 .08 .14years Low High total

D 4 1 year .15 .13 .103<D<4

no degree.04 .10

1 < D 4 2 years .27 .11 .163<D<4

degree.08 .10

First-Order SM .07

2 < D 4 3 years .11 .07 .08 SubsamplesFirst-Order .09m

3 < D 44 no degree .05 .16 .04 DurationSubsamples

3 < D 44 degree .06 .15 .08 Total Sample .06All Second-Order .13

O<D44years .04 .10 .06 Subsamples-Grand Average

TABLE F TABLE H

2 2RB+c - RB for First- and Second-Order

Subsomples and the Total Sample

Pooled Ri4c RI for First- and Second-Order

Subsamples and the Total Sample Comparedto Original Differences

Subsampledesignation

Low Original SMHigh Original SM

Original Pooled

.03 .12

.05 .11

D 4 1 .07 .13

Duration (D) of 1<0<2 .13 .21

exposure to collegiate 2<0<3 .08 .11

socializstion, in Original SM Subtotal/ 3<044 .01 .osyean Low High total no degree

D 4 1 .11 .14 .07 3<044 .04 .os

&area1 < D 4 2 .30 .11 .13 First-Order SM .0311111M.

Subsamples

2<D43 .11 .11 .08 First-Order .07

Duration3< D 44 no degree .03 .09 .01 Subsarmles

Total Sample .02 IMIN1

3 < D 44 degree .04 .08 .04 NI Second-OrderSubsamles-

1=1, .11

0<D44 .03 .05 .02 Grand Average

3935

TABLE I

R2 - R2 for First- and Second-OrderB+C CSubsamples and the Total Sample

Duration (D) ofexposure to collegiatesocialization, inyears

D < 1

1 < D < 2

2<D<3

3 <0 < 4 no degree

3 < D < 4 degree

0<D<4

Original SMLow High

Subtotal/total

.06 .17 .08

.09 .18 .07

.06 .40 .12

.07 .05 .02

.07 .04 .05

.04 .03 .03

TABLE J

Pooled R 2 - R 2 for First and Second-OrderB+C CSubsamp les and the Total Sample Compared to

Original Differences

Subsample

designation Original Pooled

Low Original SM .04 .07High Original SM .03 .17D < 1 .08 .111<D2 .07 .142<D3 .12 .233<D4

no degree

.02 .06

3<D4degree

.05 .06

First-Order SM - .04Subsamples

First-Order .07DurationSubsamples

Total Sample .03 ___

All Second-Order ........ .12Subsamples-Grand Average

.Y,'4,!,gf")43,'Vg 41'he.0 r ,ViCAWAIINVIrfIl4/11../11101.m.

References

AAUP. The economic status of the profession,1963-4. AAUP Bulletin, Summer 1964,

139-184.

Allwood, M. S. (Ed.). The new generation: Studiesin the attitudes and behavior of contemporaryAmerican students. Mt. Pleasant, lowa: lowaWesleyan College: The New Prairie, 1964.

Anderson, A. The modernization of education. InM. Weiner (Ed.), Modernization: The dynamicsof growth. New York: Basic Books, 1966.

Astin, A. W. An empirical characterization ofhigher educational institutions. Journal of Edu-cational Psychology, 1962, 53, 224-235.

Astin, A. W. Differential college effects on themotivation of talented students to obtain thePhD. Journal of Educational Psychology, 1963,54, 63-71.

Astin, A. W. Personal and environmental factorsassociated with college dropouts among highaptitude students. Journal of EducationalPsychology, 1964, 55,219-227.

Astin, A. W. Who goes Mere to college? Chicago:Science Research Associates, 1965.

Astin, A. W. The college environment Washington,D.C.: American Council on Education, 1968.

Astin, A. W., & Holland, J. L. The environmentalassessment technique: A way to measure college

environments. Journal of Educational Psychol-ogy, 1961, 52, 308-316.

Baggeley, A. R. Intermediate correlational meth-ods. New York: Wiley, 1964.

Barker, R. G., & Gump, P. V. Big whool, smell

school. Stanford, Calif.: Stanford UniversityPress, 1964.

37

Barron, F. Creative vision and expression in writingand painting. in D. W. MacKinnon (Ed.), Con-ference on 'The creative person." Berkeley:University of California, University Extension,Liberal Arts Department, 1961.

Barton, A. H. Studying the effects of collegeeducation: A methodological examination of"Changing values in college." New Haven,Conn.: Hazen Foundation, 1959.

Beach, L. R. Personality change in the church-related liberal arts college. Psychological Re-ports, 1966, 19, 1257-1258.

Beach, L. R. Study of personality in the church-related liberal arts college. Journal of CollegeStudent Personnel, 1967, 8, 105-108.

Berdie, R. F. Personality changes from high schoolentrance to college matriculation. Journal ofCounseling Psychology, 1968, 15, 376-380.

Bolton, C. D., & Kammeyer, K. C. W. Theuniversity student. New Haven, Conn.: College& University Press, 1967.

Brim, 0. G., Jr. Adult socialization. In J. A.Clausen (Ed.), Socialization and society. Boston:Little, Brown & Co., 1968.

Brown, D. G. Academic lebor markets. A report tothe office of Manpower, Automation & Train-ing, U.S. Department of Labor, September1965.

Caplow, T. The sociology of work. New York:McGraw-Hill, 1964.

Canter, A. M. (Ed.) American universities andcolleges. Washington, D.C.: American Councilan Education, 1964.

41

Cass, J., & Birnbaum, M. Comparative guide toAmerican colleges. New York: Doubleday,1964.

Cass, J., & Birnbaum, M. Comparative guide toAmerican colleges. New York: Doubleday,1966.

Clark, B. R. Educating the expert society. SanFrancisco: Chandler, 1962.

Clausen, J. A. (Ed.). Socialization and society.Boston: Little, Brown & Co., 1968. (a)

Clausen, J. A. Recent developments in socializationtheory and research. Annals of the AmericanAcademy of Political and Social Sciences, 1968,377, 139155. (b)

Coleman, J. S. The adolescent society: The sociallife of the teenager and its impact on education.New York: Free Press, 1961.

Coleman, J. S. Introduction to mathematical sod-ology. New York: Free Press, 1964.

The college blue book. Yonkers, N.Y.: C. E.Burckel, 1965.

Crutchfield, H. S. Independent thought in a con-formist world. In S. Larber & R. H. L. Wilson(Eds.), Conflict and Creativity. New York:McGrawHill, 1983.

Deutsch, M., & Krauss, R. M. Theories in socialpsychology. New York: Basic Books, 1985.

Draper, N. R., & Smith, H. Applied regreuionanalysis. New York: Wiley, 1968.

Duncan, 0. D., & Rein, A. J., Jr. Social character-istics of urban and rural communities. NewYork: Wiley, 1958.

Durkheim, E. The division of labor in society. NewYork: MacMillan, 1933.

Eckert, R. Outcome of "moral education. Minne-apolis: University of Minnesota, 1943.

Eddy, E. D., Jr. The college influences on studentcharacter. Journal of the National Association ofWomen Deans and Counselors, 1959, 22,183-186.

Editor & Publisher Co. Editor & publisher 1958market guide. New York: Author, 1957.

Eisenstadt, S. N. From generation to generation.New York: Free Press, 1958.

Elton, C. F. Patterns of change in personality testscores. Journal of Counseling Psychology, 1969,16, 95-99.

Elton, C. F., & Rose, H. A. The face of change.Journal of Counseling Psychology, 1968, 15,372-375.

Feldman, K. A. Research strategies in studyingcollege impact. ACT Research Report No. 34,Iowa City, Iowa: The American College TestingProgram, 1970.

Feldman, K. A., & Newcomb, T. M. The impact ofcollege on students. San Francisco: Jossey-Bass,1989.

Flacks, R. Adaptations of deviants in a collegecommunity. Unpublished doctoral dissertation,University of Michigan, 1983.

Freedman, .M. B. impact of college. Washington,D.C.: U.S. Department of Health, Education,and Welfare,'Office of Education, 1960.

Freedman, M. B. 7he college experience. SanFrancisco: Jossey-Bass, 1967.

Goffman, E. The presentation of self in everydaylife. Garden City, N. Y.: Doubleday, 1959.

Goldsen, R. K., Rosenbert, M., Williams, R. M., &Suchman, E. A. What college students think.Princeton, N,J.: D. Van Nostrand, 4960.

Gordon, C. W. The social system of the highsthool. New York: Free Press, 1957.

Goslin, D. A. The school in (=temporary society.Chicago: Scott-Foresman, 1985.

42

Gottlieb, D. Processes of socialization in Americangraduate schools. Social Forces, 1961, 40,124-131.

Gottlieb, D., & Hodgkins, B. J. College studentsubcultures: Their structure and characteristicsin relation to student attitude change. SchoolReview, 1963, 71, 266-289.

Gough, H. G. Techniques for identifying thecreative research scientist. In D. W. MacKinnon(Ed.), Conference on "The creative person."Berkeley: University of California Extension,Liberal Arts Department, 1961.

Haller, A. 0., & Wolff, E. W. A note on "Person-ality orientations of farm, village, and urbanboys." Rural Sociology, 1965, 30, 338-340.

Harris, C. W. (Ed.). Problems in measuring change.Madison, Wisc.: University of Wisconsin Press,1963.

Havighurst, R. J., & Taba, H. Adolescent characterand personality. New York: Wiley, 1949.

Heist, P. A. Diversity in college student character-istics. Journal of Educational Sociology, 1960,33, 279-291.

Heist, P. A., & Yonge, G. D. Omnibus personalityinventory manual. New York: The PsychologicalCorporation, 1968.

Helson, R. Creativity, sex, and mathematics. InD. W. MacKinnon (Ed.), Conference on "Thecreative person." Berkeley: University of Califor-nia Extension, Liberal Arts Department, 1961.

Herriott, R. E., & Hodgkins, B. J. Environment andeducation: A sociological analysis (title tenta-tive). Englewood Cliffs, N.J.: Prentice Hall, inprow

Hodgkins, B. J. Student subculturesAn analysisof their origins and affects on student attitudeand value change in higher education. Unrab-lished doctoral dissertation, Michigen StateUniversity, 1964.

Hodgkins, B. J., & Herriott, R. E. Age-gradestructure, goals, and compliance in the school:An organizational analysis. Sociology of Educa-tion, 1970, 42, 36-42.

Hollingshead, A. B. Elmtown's youth: The impactof social classes on adolescents. New York:Wiley, 1949.

Inkeles, A. Sociology and psychology. In S. Koch(Ed.), Psychology: A study of a science, Vol. VI.New York: McGraw-Hill, 1963.

lnkeles, A. Social structure and the socialization ofcompetence. Harvard Educational Review, 1966,36, 265-283. (a)

lnkeles, A. The modernization of man. In M.Weiner (Ed.), Modernization: The dynamics ofgrowth. New York: Basic Books, 1966. (b)

Inkeles, A. Social structure and socialization. In D.A. Goslin (Ed.), Handbook of socializationtheory and research. Chicago: Rand McNally,1969. (a)

Inkeles, A. Making men modern: On the causes andconsequences of individual change in six devel-oping countries. American Journal of Sociology,1969, 75, 206226. (b)

lnkeles, A., & Levinson, D. J. The personal systemand the sociocultural system in large-scale organ-izations. Sociometry, 1963, 26, 217-229.

Jacob, P. E. Changing values in college! Anexploratory study of the impact of generaleducation in social sciences on the values ofAmerican students. New Haven, Corm.: HazenFoundation, 1956.

Jacob, P. E. Changing values in college: Anexploratory study of the impact of collegeteaching. New York: Harper & Row, 1957.

Jahoda, M. Cument concepts of mental health.New York: Basic Books, 1959.

Jencks, C., & Riesman, D. The aced--urrik revolu-tion. Garden City, N.Y.: Doubleday, 1968.

3943

Kahl, J. A. The American class structure. NewYork: Holt, Rinehart & Winston, 1957.

Katz, J. et al. No time for youth. San Francisco:Jossey-Bass, 1968.

Klapp, 0. E. Heroes, villains and fools:The chang-ing American character. Englewood Cliffs, N.J.:Prentice-Hall, 1962.

Knapp, R. H., & Greenbaum, J. J. The youngerAmerican scholar; His collegiate origins. Chi-cago: University of Chicago Press, 1953.

Kom, H. A. Personality scale changes from thefreshman to the senior year. In J. Katz (Ed.),Growth and constraint in college students: Astudy of the varieties of psychological develop-ment. Washington, D.C.: U.S. Department ofHealth, Education, and Welfare, 1967.

Lazarsfeld, P. F., & Thielens, W., Jr. The academicmind. Glencoe, Ill.: Free Press, 1958.

Learned, W. S., & Wood, B. D. The studynt and hisknowledge. New York: Carnegie Foundation,1938.

Lehmann, I. J., & Dressel, P. L. Changes in criticalthinking ability, attitudes, end values associatedwith college attendance. U.S. Office of Educa-tion Cooperative Research Project No. 1648.East Lansing, Mich.: Michigan State University,1963.

Lehmann, I. J., & Payne, I. K. An exploration ofattitude and value changes of college freshmen.Personnel and Guidance Journal, 1963, 41,403-408.

Lerner, D. The passing of traditional society.Glencoe, Ill.: Free Press, 1958.

Lord, F. M. The messurement of growth. Eduat-tional and Psychological Measurement, 1958,16, 421-437.

Lord, F. M. Further problems in the measurementof growth. Educational and PsychologicalMeasurement, 1958, 18,437-451.

Lundberg, F. The coming world transformation.Garder City, N.Y.: Doubleday, 1963.

Mack, R. W. Transforming America: Patterns ofsocial change. New York: Random House, 1967.

MacKinnon, D. W. Creativity in architects. In D. W.MacKinnon (Ed.), Conference on '7he creativeperson." Berkeley: University of CaliforniaExtension, Liberal Arts Department, 1961.

Mannheim, K. Man and society in an age ofreconstruction. New York: Harcourt, Brace &Co., 1940.

Maslow, A. H. Toward a psychology of being.Princeton, N.J.: Van Nostrand, 1962.

Mayhew, L. B. Colleges today and tomorrow. SanFrancisco: Jossey-Bass, 1969.

McConnell, T. R. et al. Student developmentduring the college years, in press.

Mechanic, D. Students under stress. NfsW York:Free Press, 1962.

Moore, 0. K., & Anderson, A. R. Some principlesfor the design of clarifying educational environ-ments. In D. A. Goslin (Ed.), Handbook ofsocialization theory and research. Chicago: RandMcNally, 1969.

Moore, W. E. The impoct of industry. EnglewoodCliffs, N.J.: Prentice-Hall, 1965.

Newcomb, T. M. Personality and social change.New York: Dryden Press,1943.

Newcomb, T. M. The general nature of peer groupinfluence. In T. M. Newcomb & E. K. Wilson(Eds.), The study of college peer groups: Prob-lems and prospects for research. Washington,D.C.: Social Science Research Council, 1961.

Newcomb, T. M. et al. Persistence and change:Bennington College and its students aftertwenty-five years. New York: Wiley, 1967.

Newcomb, T. M., & Feldman, K. A. The impact ofcolleges upon their students. A report to theCarnegie Foundation, 1968.

44

Nichols, R. C. Effects of various college character-istics on student aptitude test scores. Journal ofEducational Psychology, 1964, 55, 45-54.

Nichols, R. C. Personality change and the college.American Educational Research Journal, 1967,4, 173-190.

Pace, C. R. They went to college. Minneapolis:University of Minnesota Press, 1941.

Pace, C. R. Five college environments. CollegeBoard Review, 1960, 41, 24-28.

Pace, C. R. The influence of academic and studentsubcultures in college and university environ-ments. Los Angeles: University of California,1964.

Palley, D. B. The aims of American colleges anduniversities. An honors thesis for the Depart-ment of Social Relations at Harvard College,1968.

Parsons, T. The social system. New York: FreePress, 1951.

Parsons, T., & Platt, G. M. The American academicprofession: A pilot study. Cambridge, Mass.:Harvard University Lithograph, 1968.

Parsons, T., & Platt, G. M. Higher education,changing socialization, and contemporary stu-dent dissent. In M. W. Riley et al. (Eds.), Asociology of age stratification. New York:Russell Sage Foundation, in press.

Pavalko, R. M., & Ullrich, A. E. Factors related toquality of schools attended by teachers. Journalof Teacher Education, 1970, 21, 498-503.

Pervin, L. A. A twenty-college study of student Xcollege interaction using TAPE (transactionalanalysis of personality and environment):Rationale, reliability, and validity. Journal ofEducational Psychology, 1967, 58. 290-302.

Plant, W. T. Personality changes associated with acollege environment. U.S. Office of Education,Cooperative Research Branch Project 348 (SAE7666), 1963.

Rand McNally. 1958 Road and reference atlas.

Rhodes, A. L., & Nam, C. B. The religious contextof educational expectations. American Socio-logical Review, 1970, 35, 253-267.

Riesman, D. The "Jacob report." American Socio-logical Review, 1958, 23, 732-738.

Riesman, D. Changing colleges and changing stu-dents. National Catholic Education AssociationBulletin, 1961, 58, 104-115.

Rogers, C. R. Client centered therapy. Boston:Houghton Mifflin, 1951.

Rogoff, N. Board member colleges: A comparativeanalysis. New York: Bureau of Applied SocialResearch, Columbia University, 1957.

Rosen, B. C., & Bates, A. P. The structure ofsocialization in graduate school. SociologicalInquiry, 1967, 37, 71-84.

Sanford, N. Self and society. New York: AthertonPress, 1966.

Schnore, L. F. The rural-urban variable: An urban-ite's perspective. Rural Sociology, 1966, 31,246-261.

Schwebel, M. Who can be educated? New York:Grove Press, 1968.

Scott, J. W., & El-Assal, M. Multiversity, universitysize, university quality and student protest: Anempirical study. American Sociological Review,1969, 34, 702-709.

Selvin, H. C. The effects of leadership. New York:Free Press, 1960.

Shibutani, T. Society and personality. EnglewoodCliffs, N.J.: Prentice-Hall, 1961.

Sirjamaki, J. Education as a social institution. In D.A. Hansen & J. E. Gerstl (Eds.), On educationsociological perspective& New York: Wiley,1967.

41 45

Snyder, B. R. How creative students fare inscience. In P. Heist (Ed.), Education for creativ-ity: A modern myth? Berkeley, Calif.: Centerfor Research and Development in Higher Educa-tion, 1967.

Stanley, J. C. A design for comparing the impact ofdifferent collegas. American Educational Re-search Journal, 1967, 4, 217-228.

Stern, G. G. Characteristics of the intellectualclimate in college environments. Harvard Educa-tional Review, 1963, 33, 5-41.

Stewart L. H. Change in personality test scoresduring college. Journal of Counseling Psychol-ogy, 1964, 11, 211-219.

Thistlethwaite, D. L. College press and studentachievement. Journal of Educational Psychol-ogy, 1959, 50, 183-191.

Thistlethwaite, D. L., & Wheeler, N. Effects ofteacher and peer subcultures upon student aspir-ations. Journal of Educational Psychology,1966, 57, 35-47.

Trent, J. W. The development of intellectualdisposition within Catholic colleges. Unpub-lished doctoral dissertation, University of Cali-fornia, Berkeley, 1964.

Trent, J. W. Catholics in college; Religious commit-ment and the intellectual life. Chicago: Univer-sity of Chicago Press, 1967.

Trent, J. W., & Medsker, L. L. Beyond high school.San Francisco: Jossey-Bass, 1968.

Turner, R. H. The social context of ambition. SanFrancisco: Chandler, 1964.

Tyler, F. T. A four-year study of personality traitsand values of a group of National Merit Scholarsand Certificate of Merit recipients. Center forthe Study of Higher Education, University ofCalifornia, Berkeley, 1963.

U.S. Bureau of the Census. A feport on du 17thdecennial census of the United States popula-tion. Washington, D.C.: U.S. Government Print-ing Office, 1952.

U.S. Bureau of the Census. County and city databook: A statistical abstract supplement, 1956.Washington, D.C.: U.S. Government PrintingOffice, 1957.

Webster, H. Changes in attitudes during college.Journal of Educational Psychology, 1958, 49,109-117.

Webster, H., Freedman, M. B., & Heist, P. A. Person-ality changes in college students. In N. Sanford(Ed.), Me American college. New York: Wiley,1962.

Werta, C. E., & Watley, D. J. Analyzing collegeeffects: Correlation vs. regression. AmericanEducational Research Journal, 1968, 5,585-598.

ACT Research Reports

This report is Number 47 in a series published by the Research and Development Division of The American College Testing

Program. The first 26 research reports have been deposited with the American Documentation Institute, AD I AuxiliaryPublications Project, Photoduplication Service, Library of Congress, Washington, D.C. 20540. Photocopies and 35 mm.microfilms are available at cost from ADI; order by ADI Document number. Advance payment is required. Make checks or

money orders payable to: Chief, Photoduplication Service, Library of Congress. Beginning with Research Report No. 27, the

reports have been deposited with the National Auxiliary Publications Service of the American Society for Information Science(NAPS), c/o CCM Information Sciences, Inc., 22 West 34th Street, New York, New York 10001. Photocopies and 35 mm.microfilms are available at cost from NAPS. Order by NAPS Document number. Advance payment is required. Printed copies($1.00) may be obtained, if available, from the Research and Development Division, The American College Testing Program,P.O. Box 168, Iowa City, Iowa 52240. A check or money order must accompany the request.

The reports since January 1969 in this series are listed below. A complete list of the reports can be obtained by writing to theResearch and Development Division, The American College Testing Program, P. O. Box 168, Iowa City, Iowa 52240.

No. 28 A Description of Graduates of No-Year Colleges, by L. L. Baird, J. M. Richards, Jr., & L. R. Shovel (NAPS No. 00306;photo, $3.00; microfilm, $1.00)

No. 29 An Empiricvl Occupational Classification Derived from a Theoryof Personality end Intended for Practice end Research,by J. L. Holland, D. R. Whitney, N. S. Cole, & J. M. Richards, Jr. (NAPS No. 00505; photo, $3.00; microfilm, $1.001

No. 30 Differential Validity in the ACT Tests, by N. S. Cole (NAPS No.00722; photo, $3.00; microfilm, $1.00)

No. 31 ilfro Is Talented? An Analysis of Achievement, by C. F. Elton, & L. R. Shovel (NAPS No. 00723; photo, $3.00;microfilm, $1.00)

No. 32 Patterns of Educational Aspiration, by L. L. Baird (NAPS No. 00920; photo, $3.00; microfilm, $1.00)

No. 33 Can Financial Need Analysis Be Simplified? by M. D. Orwig, & P. K. Jones (NAPS No. 01210; photo, $5.00; microfilm,

$3.00)

No. 34 Research Strategies in Studying College Impact, by K. A. Feldman (NAPS No. 01211; photo, $5.00; microfilm, $2.00)

No. 35 An Analysis of Spatial aonfiguration end Its Applketion to Research in Higher Educetion, by N. S. Cole, & J. W. LCole (NAPS No. 01212; photo, $5.00; microfilm, $2.00)

No. 36 Influence of Financial Need on the Vocational Development of Cbliege Students, by A. R. Vander Well (NAPS No.61440; photo, S5.20; microfilm, $2.00)

No. 37 Prectkes and Outcomes of Vocetional-TecIrWcal Education in Tecloical end Cornimmity Colleges, by T. G. Gartland, &J. F. Carmody (NAPS No. 01441; photo, $6.80; microfilm, $2.00)

No. 38 &mien Considerations In Educational Information SYstems, by M. R. Novick (NAPS No. 01442; photo, $5.00;microfilm, $2.00)

No. 39 Interactive Effects of Achievement Orientation end Teething Style on Academic Achieverhmt, by G. Domino (NAPSNo. 01443; photo, 16.00; microfilm, $2.00)

No. 40 An Analysis of the Structure of Vocational /nterens, by N. S. Cole, & G. R Hanson (NA49S No. 01444; photo, $5.00;

microfilm, $2.00)

No. 41 How Do Community College Transfer end Occupational Student Differ? by E. J. Bros, H. B. Engen, & E. J. Maxey(NAPS No. 01445; photo, $5.50; microfilm, $2.00)

No. 42 Applications of Beyesian Methods to the Predicebn ef Educationel Performence, try M. R. Novick, P. H. Jackson, D. T.

Thayer, & N. S. Cole (NAPS No. 01544; photo, $5.00; microfilm, $2.00)

No. 43 Towerd More Equfteble Distribution of CrRege Student Aid Funds: Problems in Assessing Student Financial Need, by

M. D. Orwig (NAPS No. 01543; photo. $6.00; microfilm, $2.00)

No. 44 Converting Test Data to Cbuneliog Informetkm, by Dale J. Prediger (NAPS No. not ennoble at this time.)

No. 45 The Accuracy of Self-Report Information Collected on the ACT Test Battery: Nigh School Grades end Ions ofNonacademic Achiemment, by E. J. Maxey, & V. J. Ormsby (NAPS No. not available at this time.)

No. 46 Cbrrelates of Student Interest in Social laves, by R. H. Fenske, & J. F. Carmody (NAPS No. not available at this time.)

4347