4.9:”? :59.
.,

   

 

.. . ‘ . ‘ .
fl. ,. % wflgw

                     

if

. éfk.

 

     

 

 

     

.r
l. ‘ , 1M? 51..
I: 25.1%... ‘JNMufir/ . trial metqwm...4
my.” H... r. r a: dynrflrfr 1
8

  
 

 

      

    

n. .4121. »

     

 

   

 

 

a... a:
‘
. a
r. .

 

    

   

 

 

.
x1352.)

 

 

.
f.
2

 

 

            

 

'3 .7 "Human”?
3.... $5.22.
931%;
f 4.

74
a

 

 

 

 

 

 

 

      

to

           

I; . 3’
firfifimufl
Emwamh
Emma»?
fififinflflfi;
, .. gig"?
# I.
t :31. .W . A T
u .szmwrfirflflfihg
. 41», flflrdnhui khaki“
our nfiruflfimwpmw
. .. .flrdzfinm

 

 

 

     

       

 

1

              

(0‘1!
.Zafipw
flan- In..."

 

                 

    

                       

w pr
(BL
viflurfv
i... I Mr rill...
wfiflffldrgyg . 1
(1,4,! .l- I!
4hr.
Mr, 5“

“gt-7.! 1
«Fungi/i.

ii.
.52.. (I! :5. I
{I oiplbr. 1. .

 

  

 

55.1.
VI, vii-11f )Ix’r falli¢
. / .r 92...]! J} 1;

Wflyfiiglfhu .rIJurlvna.
r: waruflrrlaiwnufl
.) 11-9: r .

 

             

               

.z k? r.
1:. 1:

    

     

(T'd.‘4’;’
315.15}
E. r tryfyrhu:
. 1n. ananuiéwfinsi
(13.2 fiJfivV/ilv Liv}!!!
5.3.15 1? . Irv/T. )ifft
- 2.11 Ti.’ '3 i’fl 3”fi'rrlu3fl

:r 1 run)”; V 11:5.

, ([13 It”?
, r . , .
Lbflfiwflwfmwmflfiflnflzt:
11: ; , . I,
(Irtivirlrrvflr

  
   

     

   

 

 

 

r 5 .. (ray/tun:
.5 s .o :5 z
liq/.1] fririfrrfiir
74.x, VIr. vilnr‘f l’v?w7.
is;

 

  

    

14.1.”..li
1.5.1.; 1...}
IYrVIvffi;
.4 (1’11 33!:
.J flflfiet'vianlf
.r 11...:
...P...ru.lfltcl!3!
..tl.l....rt:..tlri.§. .. 1.....1;
idly/1.17:)...“ (MI! (risk
1 . I. r. ,

 

 

    

               

 

 

.y «Lu! 1 1 :I.r

tr
(I!) FI/rlv
2 {1: .ffil?

 

 

i; r,....«L
(I!!!
f ..

.r.
.171” If ‘

 

     

   
 

4... »/

           

   

 

     
      

 

   

   

,../f....:..
7.1.! 4.3:} ‘ .. . a .
3.3.1.: . . . .1} 2...
2.1;: V
r‘(5 . m
‘ “1 “an. , n
‘ .5. .. .. .
‘. , ..§.A§z§amfi
.L . 3 ; , . ‘ .
. Fifi ; A 2535.4 .1“! P)... .1. w. .,
.x.n.s.fi.. «.5 a“; [T ... . .
5‘ «A . .u..) . \i .
Liza... .Cx .
.31.

   
  

L I B R A R Y
Michigan State
Univers - ty,

  
 
  

\Hi’.’93

This is to certify that the l
thesis entitled
FACTORIAL ANALYSIS OF INTELLECTUAL INTEREST \
AND MEASUREMENT OF ITS VALIDITY
IN THE PREDICTION OF COLLEGE SUCCESS \

presented by

Yung Che Kim

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Education

   
 

Major professor

Date Ma 18 1 1

0—7639

‘ llIlfllllllllflljl(llulwljjlllfllllMW“!

 

ABSTRACT

FACTORIAL ANALYSIS OF INTELLECTUAL INTEREST
AND MEASUREMENT OF ITS VALIDITY
IN THE PREDICTION OF COLLEGE SUCCESS

BY

Yung Che Kim

The purpose of the present study was to determine the
identity and structure of the psychological construct of
intellectual interest and to measure the validity of this
construct in the prediction of college success.

The population selected for the study consisted of
all freshmen entering Michigan State University in Fall,
1970, with the following exclusions: foreign students,
transfer students, students for whom data were incomplete
and students who dropped out before the end of Fall term,
1970. Five thousand four hundred and sixty-eight students
classified as Freshmen and not included in one or more of
the above exclusions registered for credit courses during
the Fall registration period. From this restricted popula—
tion, six hundred and forty—three students were randomly
selected.

The Academic Interest Scale and the M.S.U. Student

 

Survey were given to the sample of students during the

orientation week, September 21—22, 1970. The Academic

Yung Che Kim

Interest Scale included four subscales from the Stern

 

Activities Index, namely, Reflectiveness, Humanities-Social

 

Science, Understanding, and Science; the Intellectual

 

Interest Scale of Anderson and Western; and the Intellec-

 

tualism-Pragmatism Scale of Yuker and Block. The M.S.U.

 

Student Survey contained several scales. They were Trait

 

Self-Ratings of College Freshmen, the Life Goals of College

 

 

Freshmen, the General Self-Concept of Academic Ability, and

 

the Student Questionnaire. Other data such as score on the

 

Scholastic Aptitude Test, MSU Reading Test, MSU Arithmetic

 

 

 

Test, MSU Mathematics Test, high school grade point average

 

and MSU grade point average were also obtained.

The composite score of the Academic Interest Scale,

 

used to measure the construct of intellectual interest, was
derived by applying a 2-point scaling system to all of the
79 items in the three subscales, thus assigning equal value
to each of the items. The correlation coefficient of the
composite score based on the 2-point scaling system was .89
with the Stern total score, .70 with the Anderson and
Western total score, and .63 with the Yuker and Block total
score. The high intercorrelation coefficients evidenced
both the convergent validity of the three subscales of
intellectual interest and the justification of the method
of derivation of the composite score. This manner of deri-
vation of the composite score was further justified
because items in the Likert—type scales were considered to

be of approximately equal value. The KR 21 estimate of

 

 

 

Yung Che Kim

reliability was .83. It suggested that the composite scale
of intellectual interest was very reliable.

The identity and structure of the construct of intel-
lectual interest was investigated by a principle axis
solution of a 30 X 30 correlation matrix of the variables
related to the cognitive, affective and background charac-
teristics. Ten factors of acceptable magnitude, as
indicated by their eigenvalue, were included in the Varimax
rotation procedure. They were labeled scholastic aptitude,
social sensitivity, high school achievement, parents'
educational level, aesthetic, community size, conforming-
religious, scientific, social hedonism and intellectual
interest.

The three subscales of the composite scale of intel-
lectual interest and the content of the "humanistic-
cultural life goal" formed the common factor labeled
"intellectual interest."

From the critical examination of the operational
definitions drawn in the three subscales of intellectual
interest, which were highly related, and the content of the
"humanistic-cultural life goal," it was suggested that
there were three highly correlated aspects of the construct
of intellectual interest. They were an appreciation and
enjoyment of cultural pursuits, academic and philosophical
enquiry and antipragmatic interests in the arts as well as

in the sciences.

 

Yung Che Kim

The test of analysis of variance indicated that there
were statistically significant differences at d = .01,
although small in magnitude, in intellectual interest,

measured by the Academic Interest Scale, among (1) students

 

majoring in different curricula, (2) students having dif—
ferent levels of educational expectations, and (3) students
whose fathers have different occupations.

In order to investigate the validity of intellectual
interest in the prediction of college success as measured
by academic grade point average, the Pearson product—moment
correlation coefficient, r = .12, and the correlation
ration, nyx = .24, were obtained. The former was based on
the linear regression model and the latter on the curvi—
linear model. Both of the predictive validities were found
to be statistically significant, but they were not statis—
tically different. Therefore, the linear model of regres—
sion was adopted and it was concluded that the curvilinear
model would provide no better prediction of college success
than the linear model. Graphical examination of the dis-
tribution provided the same conclusion.

Also, the incremental validity was measured to test
whether or not the addition of intellectual interest to
four other commonly used predictors increased the predicta-
bility of college success. None of these increases was
statistically significant. Therefore, it was concluded that
intellectual interest had no incremental validity when

added to these predictors.

FACTORIAL ANALYSIS OF INTELLECTUAL INTEREST
AND MEASUREMENT OF ITS VALIDITY

IN THE PREDICTION OF COLLEGE SUCCESS

BY

Yung Che Kim

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Counseling, Personnel Services,
and Educational Psychology

1971

ACKNOWLEDGMENTS

I wish to express sincere appreciation to my major
advisor, Dr. James W. Costar, who has been a source of
encouragement and helpfulness throughout my doctoral
program and to Dr. Raymond N. Hatch, Dr. Norman Abeles,
and Dr. LeRoy A. Olson for their contributions and interest
as members of the writer's guidance committee.

My thanks are extended to Dr. Willard G. Warrington
Director of the Office of Evaluation Services, for pro-
viding the writer with data essential to the study, and to
Dr. Irvin J. Lehmann, Dr. Walker H. Hill and Dr. Arvo E.
Juola for their critical examination of the manuscript and
for their friendship while I was a graduate assistant in
Evaluation Services. Special thanks are extended also to
Dr. Andrew C. Porter for his assistance in the statistical
treatment of data and constructive criticism of the
manuscript.

Finally, I must recognize my indebtedness and grati-
tude to my father in memory and my mother; also to my wife,
Myung Hae Lee, daughter, Sue Im, and son, Juhn Woo, who
have had to sacrifice and patiently postpone our shared
life while we were living apart, in Korea and in America,

during the years of my graduate study.

ii

TABLE OF CONTENTS

ACKNOWLEDGEMNTS . . . . . . . . . . .
LIST OF TABLES . . . . . . . . . .
CHAPTER
I. INTRODUCTION . . . . . . .
Importance of the Study . . . .

Purpose of the Study . . . . .
Hypotheses to Be Tested . . . .
Definitions of Terms . . . .
Assumptions . . .
Limitations of the Study .
Summary of the Chapter . .

II. REVIEW OF THE LITERATURE . . . . .

Studies Relating to the Domain of
Intellectual Interest .
Studies Relating to the College and
University Influence . . . .
Prediction Studies of College Success
Studies Using Cognitive Variables
Studies Using Affective Variables
Studies Using Other Variables .
Studies Related to Some Other
Problems of Prediction . .
Summary of the Chapter .

III. METHODOLOGY . . . . .

Population and Sample . . . . .
Instrumentation and Criterion . . .
Collection of Data . . . . . .

The Statistical Models . . .
Statistical Hypotheses . . . . .
Summary of the Chapter . . . . .

iii

Page

ii

15

21
27
28
31
37

39
41

44

44
45
6O
61
67
70

IV. ANALYSIS OF THE DATA . . . . . . .

Section 1: The Results of Factor Analysis
and Analysis of Variance . . . . .

Hypotheses Tested . . .

Description of Variables Used in the
Factor Analysis Procedure . . .

Dimensions Identified Through Factor
Analysis .

The Hypotheses Testing and Definition
of the Construct of Intellectual
Interest .

Section 2. Reliability, Validity and
Regression Model . . .

Section 3: Incremental Validity of
Intellectual Interest with Cumulative
College GPA as a Criterion and Some
Cognitive and Affective Variable
Predictor(s) . . . . . . . . .

Summary of the Chapter . . . . . .

V. SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS
Overview of Purpose and Instrumentation
Data Analysis, Results and Discussion
Suggestions for Further Research . .

BIBLIOGRAPHY . . . . . . . . . . . . .

APPENDICES . . . . . . . . . . . . . .

iv

Page
72
73
73
75

79

89

97

103
111

114
114
116
127
131

143

LIST OF TABLES

 

 

Table Page
3.1 Inter—Scale Correlations of An Inventory

of Students' Attitudes by Anderson and

Western . . . . . . . . . . . . . 48
3.2 Intercorrelation Coefficients of Variables

Based on the Different Subscales and Dif—
ferent Scoring Systems of Intellectual

 

 

 

Interest Scale . . . . . . . . . . . 52
3.3 Correlations Among Promax Oblique Factors of

Trait Self-Ratings of College Freshmen . . . 55
3.4 Intercorrelations of the Factors of the

Life Goals of College Freshmen for the

Present Sample . . . . . . . . . . . 57
4.1 Mean Score and Standard Deviation of Each of

30 Variables Used in the Factorial Analysis . 78
4.2 Eigenvalues of 30 Factors Extracted Through

Principle Axis Method . . . . . . . . 79
4.3 Varimax Rotation Analysis—~Rotated 10

Factors with Rounded Loadings for 30

Variables . . . . . . . . . . . . 81
4.4 Varimax Rotated 3O Variables Loading Above

.50 on 10 Factors Which Provided the

Factor Interpretation . . . . . . . . 83
4.5 Variable Content of Factor l——Scholastic

Aptitude . . . . . . . . . . . . . 84
4.6 Variable Content of Factor 2--Social

Sensitivity . . . . . . . . . . . . 85
4.7 Variable Content of Factor 3—-High School

Achievement . . . . . . . . . . . . 85
4.8 Variable Content of Factor 4--Parents'

Educational Level . . . . . . . . . . 86

v

4.16

4.18

Variable Content of Factor 5-—Aesthetic. . .

Variable Content of Factor 6--Community
Size . . . . . . . . . . . . .

Variable Content of Factor 7--Conforming-
Religious . . . . . . . . . . . .

Variable Content of Factor 8——Scientific

Variable Content of Factor 9--Social
Hedonism . . . . . . . . .

Variable Content of Factor lO--Intellectual
Interest . . . . . . . . . . . .

Mean Score and Standard Deviation of
Intellectual Interest Test Score on Ten
Categories of Major Fields .

Analysis of Variance of the Variable of
"Major Field" with the Dependent Variable
of Intellectual Interest . . .

Mean Score and Standard Deviation of
Intellectual Interest Test Score on Six
Categories of Educational Expectation

Analysis of Variance of the Variable of
"Educational Expectation" with the
Dependent Variable of "Intellectual
Interest" . . . . . . . .

Mean Score and Standard Deviation of
Intellectual Interest Test Score on Nine
Categories of Father's Occupation .

Analysis of Variance of the Variable of
"Father's Occupation" with the Dependent
Variable of "Intellectual Interest" Score

Analysis of Variance for Overall Regression
of Intellectual Interest with Cumulative
College GPA as a Criterion Variable .

Regression Coefficient and Its Standard
Errors and Standardized Beta Weights and

Its Standard Errors of Intellectual Interest
with Cumulative College GPA as a Criterion.

vi

Page

86

87

87

88

88

89

93

93

94

95

96

96

99

100

 

Table

4.25

4.26

Scatter Diagram of Intellectual Interest
Score, Based on 24 Categories, on
College GPA . . . . . . . . . . .

Mean Score and Standard Deviation of
Criterion Variable of MSU GPA and Five
Predictors Used in the Multiple Regression
Equation . . . . . . . . . . . .

Intercorrelation Coefficient of Criterion
Variable of MSU GPA and the Five Predictors
Used in the Multiple Regression Equation

Simple and Multiple Correlation Coefficients

and Constants and Regression Coefficients
of Simple and Multiple Regression Equations

vii'

Page

104

107

107

109

 

CHAPTER I

INTRODUCTION

Ability to think is the central concern of a univer-
sity because it is a primary factor in the learning pro-
cess. To say that the ability of students to think

clearly is of central concern is neither to say that it is

 

the sole interest of the university nor, in all circum—
stances, the most important, but that it is a pervasive
concern in all aspects of the work of a university.

With an increasing demand for intellectual compe-
tence in modern times, larger numbers of youth from all
segments of our population are entering colleges and
universities. More and more young people and their
parents are viewing college education as necessary in
order to achieve success in a complex society.

Some of the benefits of higher education include a
stronger belief in one's own capacities to handle broad
responsibilities, an increased ability to solve important
problems, and a greater likelihood of making a significant
impact on the larger society. Higher education is an
important aspect of the continuing process of rationaliza-
tion in all Spheres of life-~finding logical and coherent

patterns in the flux of events.

1

IIIIIIIIIIIIIIIIIIIIIIIIII--______________________________________

 

Faced with limited budgets and physical facilities,
institutions of higher education are finding it increas-
ingly difficult to meet the diverse demands of students,
their parents and society in general. Rapid psychological
and technological advances in recent years have increased
the demand upon educators to "toughen it up, pour it on,
and raise the standards of excellence."

Rising standards of excellence have, in turn, caused
colleges and universities to become more selective in
their admission procedures. Differences in admission
requirements and the great variation in the content of the
curricula found among our colleges and universities today
are factors which help build the uniqueness of each insti—
tution's educational environment. These differences among
universities along with those resulting from a wide range
of abilities within the student body and great diversity
among the high school academic training programs from
which students come, are bewildering to students. Conse-
quently, institutions of higher education must, on an
individual or group basis, develop a policy for admission
which includes matching the academic style of the student
with the demands of the scholarly environment that exists
on the campus.

At present, the selection of students is determined
by many factors, not all of which are easily identified
or measured. Many research studies have indicated that

cognitive measures account for 30 to 45 percent of the

 

 

variation in academic performance. While no other single
factor accounts for this much variation, more than half of
the variance still remains unaccounted for.

From time to time, assessment of the affective
domain, including attitude and personality characteristics,
has been used as a means of describing variance otherwise
unaccounted for in academic success. It is in this area
that it seems most likely at the present time that addi—
tional research will provide information to institutions
of higher learning useful in selecting those students who
will learn with the greatest efficiency and economy.
Findings in this area will also have value for the indivi—
dual student in developing self-understanding, devising
educational and vocational plans, and making personal and

social adjustments.

Importance of the Study

 

There is considerable research evidence to indicate
that achievement stems not only from functional capacity
but also from acquired habits, interests, attitudes and
motivation. The search for better predictors of college
success has been pursued with increased effort during the
past two decades. This has produced many prediction
studies relating to college success. Usually, an intellec-
tual measure such as a scholastic aptitude test or high
school grade point average is found to be the best single
predictor of college success, especially in the freshman

year (Lutz, 1968).

 

 

Most studies of affective measures examine the rela-
tionship of the affective domain to the achievement of
cognitive objectives. Some of the affective variables
related to these studies are individual aspiration to
succeed (Nason, 1954); motivation to achieve generally.
(Morgan, 1952); self-confidence, self—acceptance, positive
self-concept (Terman, 1947); and belief in oneself (Bishton,
1966). McConnel and his associates (1962) studied "intel—
lectual potentiality" and its promise of academic success
at the Center for the Study of Higher Education, University
of California.

These studies clearly indicate that it is still pro-
ductive to study the personal characteristics of college
students as they relate to intellectual or scholastic per-
formance. The relationship of such performance to life
goals, self-ratings of traits and other cognitive and non-
cognitive variables merits further attention.

Since it is true that each individual is different,
differentially qualified and differentially characterized,
individual differences among students should be identified
and their implications for scholastic achievement investi-
gated further. Everyone, whether planning to attend
college or not, should fully understand his strengths and
weaknesses in order to cope with the academic and voca-
tional requirements of living in modern society. The
capacity to accurately measure a clearly defined construct

such as intellectual interest will help students with

 

decision-making in all areas including selection of a
college to attend, major field to study and social activi-
ties in which to engage.

In addition, institutions of higher learning are
always searching for a more valid method of screening appli-
cants for admission. It may well be that knowledge about
certain affective characteristics of prospective students
can be of major assistance both in the selection process
and in the development of effective instructional and stu—

dent personnel programs.

Purpose of the Study

 

It is the purpose of this study, first, to identify
the structure and pattern of the construct of intellectual
interest and, secondly, to test the validity of the con-
struct in the prediction of academic success of college
freshmen in an institution of higher education.

The relationship of intellectual interest to other
biographic, demographic, cognitive and affective variables
will be determined. Knowledge of these relationships may
help to describe the characteristic of intellectual
interest of college freshmen. The constitutive structure
and factorial clusters of intellectual interest are expected
to be clarified. In other words, by analyzing the construct
of intellectual interest and relating demographic, cognitive
and affective variables gathered for the study, it is pre-
sumed that the identity and structure of intellectual

interest domain will be clarified.

 

For this purpose, the following variables will be
used in the present study:

A. The biographical and demographic variables.

1. Sex.

2. Major areas the student expects to study in.

3. Community the student lives most of his/her
life.

4. Size of graduating class of high school.

5. Father's educational level.

6. Mother's educational level.

7. Educational expectation.

8. Father's occupation.

B. The cognitive variables.

1. MSU Reading test.

2. MSU Arithmetic test.

3. MSU Mathematics test.

4. MSU Quantitative test.

5. The Scholastic Aptitude test—~Verbal.

6. The Scholastic Aptitude test--Mathematics.

7. The Scholastic Aptitude test--Total.

8. Self—reported high school grade point average.

9. Actual high school grade point average.

C. The affective variables.

1. Seven factors of trait self—ratings of
college freshmen.

2. Seven factors of life goals of college
freshmen.

3. General self-concept of academic ability.

As noted in a previous section, the study of academic
predictors for college success has been popular for decades
with increasing effort and attention being given to this
area. High school grade point average, rank in class and
aptitude tests have usually been shown to be the best pre-
dictors. Such studies have revealed that aptitude tests
have an average correlation of about .50 with college grade
point average. However, none of the single variables or

combinations of variables has been successful in explaining

more than 30 to 45 percent of variation of the criterion
of academic performance. Thus, there has been an increas-
ing interest in studies focused upon affective variables.
Even with the addition of affective predictors, a large
part of the variance in the criterion of academic perfor—
mance is usually still unaccounted for.

With these factors in mind, this study is designed
to investigate and describe the relationship of the affec-
tive construct called intellectual interest to academic
performance. The study will also attempt to determine the
usefulness of intellectual interest as a predictor of

college academic success during the freshman year.

Hypotheses to Be Tested

 

The first aspect of this study will be to determine
the nature and structure of intellectual interest. For
this purpose, the following general hypotheses will be
tested:
Hypothesis 1: Three subscales used to measure the construct
of intellectual interest have convergent

validity as shown by reasonably high corre-
lations with each other.

 

Hypothesis 2: There are differences in intellectual
interest, as measured by the Academic
Interest Scale, between males and females
and among:

 

 

a. students majoring in different curricula,

b. students who lived a major portion of
their lives on a farm, in a village,
town, small city, or large city,

c. students who were in different sized
high school graduating classes,

Hypgthesis

 

Hypgthesis

 

Hypothesis

 

Hypothesis

 

Hypothesis

 

d. students whose fathers and/or mothers
completed grade school, high school,
college, graduate or professional school,

e. students who plan to receive four years
of college education and those who plan
to attend graduate or professional
school,

f. students whose fathers are executives,
business owners, white-collar workers,
skilled craftsmen, semi-skilled workers,
low or unskilled laborers, farm owners,
public service workers, or professional
personnel (doctor, lawyer, dentist, and
so forth).

There is a relationship between intellectual
interest and performance on the Scholastic
Aptitude Test, Michigan State University
Reading Test, Michigan State University
Arithmetic Test and Michigan State Univer-
sity Quantitative Test.

 

 

 

 

 

 

 

There is a relationship between intellectual
interest and seven factors of trait self—
rating measured by Trait Self—Ratings of
College Freshmen Scale.

 

 

There is a relationship between intellectual
interest and seven factors of life goals of
college freshmen measured by Life Goals of
College Freshmen Scale.

 

 

There is a relationship between intellectual
interest and academic self—concept measured
by General Self—Concept of Academic Ability.

 

The biographic, demographic and background
variables, cognitive and affective variables,
including the intellectual interest variable
in this study, may be factored into inter—
pretable subgroups.

The second aspect of the study will be to test the

validity of intellectual interest in the prediction of

academic success of freshmen students at an institution of

higher education. The following hypotheses will be tested:

Hypothesis 1: There is a significant relationship between
intellectual interest as measured by the
Academic Interest Scale and the grade point
average subjects earn during their first
term of college work.

 

 

Hypothesis 2: The accuracy of prediction of college suc-
cess is increased when intellectual interest
is considered in conjunction with a scholas-
tic aptitude test score, high school grade
point average, general self-concept of
academic ability or a combination of these
variables.

 

Definition of Terms

 

1. College grade point average (MSU GPA) refers to

 

an average of the student's grades earned in subjects com—
pleted in college work.

2. High school_grade point average (HS GPA) refers

 

to an average of all grades earned in high school subjects
and reported to the Registrar's Office of the university.

3. Predictor refers to an independent or antecedent

 

variable that provides information for forecasting an
unobserved event. The changes or differences in the pre-
dictor variable are associated with changes or differences
in the unobserved event. Values of a predictor variable
thus afford a basis for prediction of the unobserved event.

4. Criterion refers to a dependent or consequent

 

variable which is presumed to be predictable from the pre—
dictor variable or variables. A set of observable activi—
ties or behaviors that are relevant to the criterion and
that can also be measured may be called the "criterion

performance." The scores obtained on an instrument or

10

scale representing the criterion variable are termed
"criterion measures."

5. Self—concept of ability refers to the evaluation

 

a person makes of his ability to achieve in academic tasks
in general, especially as compared with others.

6. Trait self-rating refers to the evaluation a

 

person makes of himself concerning seven traits, namely,
physical well-being, scholarship, estheticism, pragmatism,
technical-scientific ability, sociability, and sensitivity
to others.

7. Life goals refers to the self-assessment of goals

 

which are most relevant for various occupations and for
various types of achievement, including seven dimensions,
namely, prestige, artistic, personal happiness, humanistic-
cultural, religious, scientific and hedonistic.

8. Regression eguation refers to the functional form

 

of the relationship between the predictors and the crite—
rion. This is expressed in the form of a mathematical
function in which Y, the criterion, is set equal to some
expression which contains values on Xs, the predictors,

and certain constants or parameters.

Assumptions

 

The predictive aspect of the study has several basic
assumptions. One of them is the assumption of linearity
of relationship between criterion and predictor. In a word,

this assumption means that, if two variables are plotted

11

one against the other, the plot tends to follow a straight
line.

Other essential assumptions regarding the use of a
regression equation are those of normality and homogeneity
of variance. Assumption of normality implies that the
samples with which we work have been drawn from populations
that are normally distributed. If the populations from
which samples are drawn are not normally distributed, then
statistical tests that depend upon the normality assumption
are violated. The homogeneity of variance assumption
implies that the variance within the groups is statisti-
cally the same. That is, variances are assumed to be
homogeneous from group to group, within the bounds of ran-
dom variation.

The assumption of linearity is tested in this study.
This assures that the assumption regarding linearity will
impose no problem for the study. Neither is there any
reason to believe that this study violated any of the other
assumptions. However, consideration will be given to these
assumptions when generalizations are drawn from the results

of the study.

Limitations of the Study

 

The manner in which the sample was selected placed
some restriction upon the external validity of the study.
Subjects who dropped out before completing the first term,
transfer students, students with incomplete data, and

foreign students were not included in the pOpulation

12

studied. Also, the population chosen for study included
only Michigan State University students. Thus, generaliza-
tions cannot be made regarding other college and university
populations. The usefulness of the prediction rule for a
certain group is dependent upon its similarity to the norm
group from which the regression equation was formulated.
To the extent that the groups are different, the prediction
will be in error.

For the study to be of maximum value it is necessary

that replicative studies be made using other populations.

Summary of the Chapter

 

The purpose of this study is to investigate the
structure of intellectual interest and to test the predic-
tive validity of intellectual interest as a predictor of
academic success in an institution of higher learning.

Nine specific hypotheses are formulated, and they will be
tested for significance.

After defining the terms frequently used throughout
the study, the need for and rationale of the study were
described from three different points of View, those of the
individual student, institutions of higher education and
society at large. Basic assumptions were then described.

Finally, limitations of the study were defined. It
was pointed out that sources of limitations were in the
theoretical assumptions of the method employed in the study

and in the restricted definition of the population.

 

13

Chapter II will present a review of related research
studies. The methodology of the study including the instru—
mentation, research procedures and statistical models
employed will be shown in Chapter III. A presentation and
analysis of the data will be conducted in Chapter IV.
Finally, Chapter V will include a summary of the study,

conclusions and recommendations for further study.

 

CHAPTER II

REVIEW OF THE LITERATURE

This study evolved from the ideas and findings of
earlier research on intellectual interest, in general, and
on college life, campus environment, and academic predic-
tion.

There has been speculation about the impact of stu-
dents upon one another and of the role of colleges and
universities as socializing institutions. The intellectual
dimensions of environmental pressure, influence of academic
emphasis on curricular activities, and the scholastic value
system were included in many previous studies. However, it
is surprising to see that little effort has been made to
directly investigate and measure these related dimensions.
Some exceptions were the studies done by Yuker and Block
(1969), Anderson and Western (1966) and McConnel and his
associates (1962).

Nevertheless, the use of tests in the selection of
applicants for admission and for the prediction of academic
success, defined in terms of college grades, has been the
most widely explored topic in educational-psychologica1
research. Segal (1934) summarized the findings of 23

studies before 1933. Garrett, in his 1949 review which

14

 

15

covered two decades, mentioned approximately 194 studies.
Fishman (1958) reported 580 studies in the years 1950-1958.
Travers (1959) in a study on the prediction of achievement,
cited more than 200 prediction related studies. Whitla
(1969) reported that the published studies exploring the
prediction relationship may represent only a fraction of
the total number of such studies that have been conducted.

The literature which has bearing on this thesis is
reviewed in this chapter. In order to put the research
reviews in perspective, the summaries of studies illustrate
the following areas:

1. Studies relating to the domain of intellectual

interest.

2. Studies relating to college and university
influence.

3. Prediction studies of college success.

a. Studies using cognitive variables.

b. Studies using affective variables.

0. Studies using other variables.

d. Studies related to other problems of
prediction.

1. Studies Relating to the Domain
of Intellectual Interest

 

 

Before examining the research related to the main
subject of the study, intellectual interest, it might be
helpful to review the main studies of the relationship

between the cognitive and affective factors associated

 

16

with learning. In the report of his international study

of educational achievement, Bloom (1965) stated that
patterns of educational objectives for a group of schools
are related to the pattern of scores on achievement exami—
nations. By that, he suggested a direct relationship
between the aims of education and academic performance as
an educational outcome. The notion that educational objec-
tives, while related to intellectual achievement, determine
the development of interests, attitudes, values, motivation
and many other non-intellectual dimensions is generally
accepted. Many scholars have documented the relationship
between the cognitive and affective domains (Scheere, 1965;
Asch, 1952; Rhine, 1958; Festinger, 1957; and Heider, 1968).
They believe in the fundamental unity of affective and
cognitive behavior. For example, Rokeach (1960) stated the
issue this way:

We assume that every affective state also has its
representation as a cognitive state in the form of some
belief or some structural relation among beliefs within
a system. With respect to the enjoyment of music, for
example, we all build up through past experience a set
of beliefs or expectancies about what constitutes
"good" or "bad" music. It is in terms of such expec—
tancies, which are more often implicit than explicit,
that we enjoy a particular composition. Thus, a person
who is exposed to a particular piece of classical music
or jazz may enjoy it, even though it may be totally
unfamiliar to him, because it is congruent with an
already existing set of beliefs he has built up over
time. Depending on the extent to which he is prepared
to entertain new systems, he may or may not enjoy
Schdnberg or other music perceived as incompatible with
his own beliefs about what constitutes good music.

. . . In all cases, enjoyment or its opposite is the
affective counterpart of a belief organization and can
be thought of as being in one-to-one relation (iso-
morphic) with it. Thus, our cognitive approach is as
much concerned with affection as with cognition.

(P. 399.)

 

l7

Rosenberg (1956), also, saw the basic congruence of
the cognitive and affective systems. Krathwohl gg_al.
(1965), claiming that it is possible that a different affec—
tive objective accompanies every cognitive objective in a
course, developed the subcategories of the two domains.

They are: knowledge-receiving, comprehension-responding,
application-valuing, analysis—conceptualization, and
evaluation-organization. Describing the categories of the
classification scheme of the affective domain taxonomy,
Krathwohl ep_gl, defined "Responding," which is the second

subdivision in the classification, this way:

 

This is the category that many teachers will find best
describes their "interest" objectives. Most commonly
we use the term to indicate the desire that a child
become sufficiently involved in or committed to a
subject, phenomenon, or activity that he will seek it
out and gain satisfaction from working with it or
engaging in it. (Pp. 118-119.)

They further defined "Valuing," the third subdivision,
in the following words:

An important element of behavior characterized by
Valuing is that it is motivated, not by the desire to
comply or obey, but by the individual's commitment to
the underlying value guiding the behavior. In the
socialization process, the learner may conform exter-
nally to a number of socially desirable rules of
behavior which he has only partially accepted as his
own--has only partially internalized. (P. 140.)

The theoretical implication, that the individual
begins to respond to stimuli and then continues increas-
ingly of his own volition to the point where he is actively
seeking instances in which he may respond, is fully

employed in forming the definition of "intellectual

interest" in the present study.

18

Guilford (1962) stated that having the information
is not sufficient in developing creativity, and Holland
(1966) suggests that creative performance occurs more
frequently among students who are intellectual and inde-
pendent.

Ellis (1962) believed that thought and emotion are
not separate or different functions. He made the following
statement:

Emotion, then, does not exist in its own right, as a
special and almost mystical sort of entity; it is,
rather, an essential part of an entire sensing-moving—
thinking-emoting complex. What we usually label as
thinking is a relatively calm and dispassionate
appraisal (or organized perception) of a given situa-
tion, an objective comparison of many of the elements

in this situation, and a coming to some conclusion as
a result of this comparing or discriminating process.

And what we usually label as emotion, . . . is a
relatively uncalm, passionate, and strong evaluating
of some person or object. (P. 47.)

Jacob (1957) suggested that affective behaviors
develop when appropriate learning experiences are provided
for students much the same as cognitive behaviors develop
from appropriate learning experiences. There is need for
conclusive experimentation and research on the relationship
between the two domains.

Furthermore, not only does the relationship between
these two domains need describing more in detail, but many
of the psychological traits need a clearer definition of
their identity and structure. This situation is also true
of the domain of intellectual interest. England and England

(1958) define an intellectual as:

. . . a person interested in ideas in contrast with the
merely practical man . . . it may be applied to one whose
interest in ideas is not balanced by practicality. (P. 267.)

19

Barnhart (1951) in the American College Dictionary
defines an intellectual as:
A member of a class or group professing, or supposed
to possess enlightened judgment and opinions with
respect to public or political questions. (P. 348.)
William James (1907) noted that the intellectualism-

pragmatism dichotomy was similar to the tender—minded,

tough—minded dichotomy. The manual of the Omnibus Person-

 

ality Inventory says:

 

The development of measures of intellectual, scholarly
concerns was channeled by what were assumed to be the
major modes or correlates of academic activity. This
was also an area where the measurement of change was
obviously related to college objectives and achieve-
ment. . . . The system of Intellectual Disposition
Categories is a way of classifying or locating persons
at certain points on a "continuum of intellectual
disposition." Specifically, the subjects are placed
in one of eight Intellectual Disposition Categories
(IDCs). This system was developed over several years
on an exploratory basis and gradually acquired suppor-
tive evidence for its validity as it was tested,
expanded, and retested. (Heist, 1968, pp. 1-56.)

Several subscales such as Thinking Introversion,
Theoretical Orientation, Estheticism and Complexity were
used by McConnel and his associates (1962) at the Center
for the Study of Higher Education, University of California.
Their emphasis was on students' intellectual interests and
dispositions at the time college students entered the
college and what happened to these characteristics while
they were in college. In their sample of 372 students in
four colleges under study, the correlations of the scales

with the Scholastic Aptitude Test—-Verbal ranged from -.01

 

to .58 and with SAT-M, from -.01 to .43. Their final study

report is not yet available.

20

Weissman (1959), using some subscales of Omnibus
Personality Inventory, made a pattern analysis of 900
National Merit Scholarship freshmen. He differentiated
three groups: Group A-—the individual tends to be thought-
ful, reflective, independent and creative; Group B——the
individual tends to have only a moderate liking for abstrac—
tions and reflective thought and Group C——the individual
tends to prefer physical activity to thoughtful reflection.

McBee and Duke (1960), after using the Brown-Holtzman

 

Survey of Study Habits and Attitudes in their study on the

 

relationship between intelligence, scholastic motivation
and academic achievement, pointed out the desirability of
the study of specific interests in determining scholastic
motivation.

Stern (1963b, 1966) investigated the college environ-
ment. Intellectual interest was one subscale of the
environmental index he developed which was found to be
significantly different from campus to campus.

One of the most extensive studies was done by Yuker
and Block. They developed a thirty-item Likert-type scale
which has been administered to over 3,500 college students.
For a sample of 134 evening students, the correlation
coefficient between grade point average and the score on
the scale they develOped was .32. When the 14 freshmen in
the sample were dropped from the statistical analysis,
grade point average for evening students classified as
sophomores and above correlated .56 with the scale. They

stated:

21

Thus, the correlation between I-P (Intellectualism-
Pragmatism Scale) score and grade point average seems
quite substantial and approaches those found between
grades and standardized tests of intellectual ability.
This is particularly interesting in View of the fact
that intelligence tests seem to have reached the point
where higher correlations with success in school do

not seem possible. . . . It is possible that the I-P
scale taps an important additional attitude related to
success in college. (P. 9.)

The I-P Scale was also administered to students who
were accepted as incoming freshmen for the Fall, 1962
class at Hostra University. For the 445 students for whom
complete data were available, the Scale correlated .26
with grades at the end of the first semester. It corre—
lated .12 with high school average. The correlations
between the I—P Scale score and each of the three GRE
scores respectively were: GRE Social Science .12, GRE
Humanities .56 and GRE Natural Science -.17. For a sample
of about 1,500 students, a linear relationship was found
between school year and the Scale score. The mean score
increased each year from entering freshmen to graduate
students. Most of the differences between two—year periods
were significant beyond the .05 level of significance.

2. Studies Relating to College and
University Influence

 

 

Various functions that peer groups serve for indivi—
dual college students have been studied extensively. An
American Council on Education research report (Astin, 1963
and 1965) gave evidence of such peer influences. Levine
(1966) and Sanford (1963) have demonstrated that the indi-

vidual's achieving independence from home was one of the

22

most important influences, and several investigators have
noted the role the peer group plays in offering general
emotional support to students, fulfilling needs not suf-
ficiently met by the faculty or classroom. Reporting on
Harvard's 1964 and 1965 classes from longitudinal data and
intensive interviews, King (1967) noted that seniors rated
the finding of meaning, goals and outlook for life as
"most important." These same students believed that their
interaction with other students was very valuable to their
maturing college experience. A Bennington study (Newcomb,
1967) showed that students bringing diverse cultural out-
looks to the college tended to be assimilated into the
cultural outlook which predominated at the college.

Berdie (1966) makes the point in the following words:
Students come from families, high schools, and com—
munities that share many of the values of the univer-
sity but that also are unaware of or perhaps rejecting
other values. From the college, the student moves
into a world of work, family, and community that again
in many ways is different from his alma mater. While
in college, the stresses and demands of the curriculum
and college life are balanced against those of social
problems, religious conflicts, racial discrimination,
civil rights, and a society out of joint. (P. 132.)

As indicated in Berdie's remarks, student culture

may be regarded as a homogeneous culture for certain pur-
poses. Also, it might be Viewed as a plurality of heter-
ogenous subgroups valuing different interests and rewarding
different activities. Interest regarding the study of
college students' attitudes, critical thinking and values

has generated many longitudinal and cross—sectional studies.

Major studies have been or are being conducted at Michigan

 

23

State (Lehmann and Dressel, 1962), Stanford and Berkeley
(Katz and associates, 1967), Cornell (Goldsen, 1960) and
Harvard (King, 1967).

Pace (1967), Stern (1963a, 1966) and others exten-
sively studied the college environment, using the College

Characteristics Index. The Index consists of three hundred

 

statements about college life—-rules and regulations,
features and facilities, faculty, curriculum, instruction,

extracurricular programs and others. The Environment

 

Assessment Technique, developed by Astin and Holland (1961),

 

also assesses the college campus culture. Centra (1967)

reworded the College and University Environment Scales and

 

applied it to dormitories rather than to the university in
general. After measuring the environmental pressures,
demand and opportunities of residence halls at Michigan
State University, Olson (1964) discovered that these living
quarters differed from another primarily along an "intellec—
tual—propriety dimension." Trow (1965) has explained it
this way:
Most colleges are not monolithic and uniform, but con—
tain within themselves different subsocieties whose
members share common codes of values, attitudes, and
patterns of behavior. . . . The kind of subcultures a
student identifies with shapes the kinds of people he
spends his time with and the kinds of values and atti—
tudes he is exposed, indeed, subjected to. (P. 58.)
Trow emphasized the interplay between personal and
institutional distinctions in the shaping of self-concepts.

Furthermore, he explained that certain attitudes, behaviors

and styles of thought and action among adults who have been

24

to college, which are of importance for the quality of life
in the society, may reasonably be believed to have been
affected by some aspect of their experience in higher edu-
cation. Moreover, Havice (1966), Bolton and Kammeyer
(1967), Siegel (1968), and Krech (1962) explored the
importance of knowledge of students' intelligence, needs,
values, pressures and other characteristics for the indi-
vidual who is on campus or going to apply for admission and
for institutional authorities as well.

Tyler (1962) noted the point as follows:

Let us consider the problem first from the standpoint
of the individual high school student. The most impor—
tant relevant questions to which he and his parents
need answers are: "In which colleges am I likely to
make the most educational progress if I continue with
my present habits and attitudes? What changes in my
habits and attitudes would be likely to result in
marked increases in my educational achievement in each
of the colleges under consideration?" . . . Information,
. . will be required to derive maximum benefit from
the college experience. . . . Let us now consider the
admission problem from the standpoint of the individual
college. The most important relevant questions to
which the college needs answers are: "What aggregations
of students are likely to make most educational pro—
gress if the college continues with its present faculty,
facilities and practices? What changes in conditions
in the college are likely to increase student learning
and for what aggregations of students?" These ques—
tions imply two assumptions which are not commonly
expressed in admission procedures. The first is that
the amount students learn is affected by the composi—
tion of the student body so that the problem of selec-
tion must consider the aggregate student body as well
as each student. . . . A second assumption is that
some of the conditions in a college affect the amount
of learning for some aggregates of students. (Pp. 106-109)

In the fall of 1964, as part of the test battery of
the American College Testing Program, over 8,000 high

school students throughout the United States who applied to

25

colleges were asked to rate influences affecting their
choices of particular colleges (Newcomb e:_gl., 1967).
Factor analysis showed the influence of intellectual empha—
sis, practicality, advice of others, social emphasis,
emphasis on religious and ethical values and size of the
school. Moreover, the composition of the entering student
body was found to be determined by the college's particular
image (Clark, 1959) or perception of the college image
(Silber, 1961). Many other studies documented the inter—
play of the individual and the institution.

In a longitudinal study of freshmen at the beginning
and seniors at the end of the school year, Trent (1967)
found that, for each sex, seniors have more liking for
reflective thought, particularly of an abstract nature,
than do freshmen. In the same type of study, Korn (1967),
Nicholas (1967), and King (1967) document the increase in
percentage of students emphasizing academic and intellec-
tual satisfactions.

Stern (1966), Brewer (1963), and Yuker and Block
(1967) have done cross—sectional studies on changes of
intellectual orientation in college. Not all of the
studies showed significant increases in intellectual orien-
tation. A set of findings by Katz (1967) showed no signi-
ficant increases. In the early 1960's, 73 percent of a
sample of women at Bennington College spontaneously men—
tioned increases in intellectuality when asked how they had

changed since coming to the college (Newcomb, 1967). In

 

26

1962, a sample of seniors at Michigan State University

were presented with a list of "behavior traits" and were
asked to describe changes that had come about in college

in terms of whether they possessed more, less or the same
degree of each quality. To the item, "interest in intel-
lectual and cultural matters," 73 percent of the men and

84 percent of the women indicated more interest (Lehmann
and Dressel, 1962). Many other studies in the general area
of intellectual orientation usually have shown changes

toward independence of thought, originality and widening

 

interests.

As socializing institutions, colleges and universi-
ties have the task of influencing students so that they
leave the campus with improved and desirable knowledge,
skills, attitudes and values. A typology of student sub-
groupings that has become popular in recent years is that
offered by Clark and Trow (1966). They describe four types
of student subcultures which they label academic, noncon—
formist, collegiate and vocational. Whealer (1966) has
analyzed the kinds of interpersonal settings and Wallace
(1966) goes so far as to caricature the interaction between
students and faculty. Astin (1963), in studying a sample
of high-ability students at some 76 different colleges and
universities, found that the intelligence level of the
student body as a whole was negatively associated with per-

ceived change in abilities and self—confidence.

27

3. Prediction Studies of
College Success

 

 

Academic maladaptiveness is one of the major problems
confronting teachers, school administrators, counselors and
students as well. An extraordinary number of studies have
been conducted on various aspects of student achievement in
America's several levels of educational institutions.

Cattel et a1. (1962) stated:

The prediction of school achievement is valuable not
only for the sheer understanding which we thereby
achieve of the psychological mechanisms and situational
conditions which lead to scholastic success, but also
for two immediate practical purposes in school organ—
ization. In the first place one wishes to discover the
causes and remedies of backgrounds in that minority of
individuals who so markedly fail to achieve that
special class organization has to be introduced. In
the second place, one may wish to select, in general
scholarship and fellowship selection practice, those
individuals who are most likely to benefit from being
given special opportunities in advanced education.

(Po 3.)

It is evident that academic failure is both a problem
to the individual, who may suffer from the sense of failure,
and to society, which loses the full potential contribu-
tions of an unestimated number of its members. It follows
that anything that can be done to reduce the incidence of
academic maladjustment will contribute to individual and
social accomplishment and well-being.

The seriousness of the problem has resulted in a
tremendous number of prediction studies, as previously
mentioned. For example, Fishman and Pasanella (1960)

report 580 research studies between 1949 and 1959 alone.

Since both the College Entrance Examination Board and the

28

American College Testing Program now provide research ser-
vices for member institutions, many colleges and universi-
ties have been able to carry out prediction studies with a
relatively small investment in terms of time, effort or
personnel. Thus, it would seem reasonable to project that
the last decade has at least equalled and likely surpassed
this figure.

The generally accepted manner of handling the liter—
ature reviews is by subdividing the broad categories of
predictors and criterion and by dealing with either intel—
lective characteristics or nonintellective characteristics
of individuals (Fishman and Pasanella, 1960). Studies
related to other problems of prediction are cited below in

four categories.

A. Studies Using Cognitive Variables

High school scholarship has been found to be the best
single predictor of college success (Beatley, 1922; Garret,
1949; Richards and Lutz, 1968). Guisti (1964) reviewed the
prediction literature and found convincing evidence that
high school grade point average was the best single pre-
dictor of college grade point average (GPA). The range of
reported correlations was .35 to .69 with a median correla-
tion of about .50. This correlation of .50, however, only
accounts for about one-quarter of the total variance of
college grade point average, indicating that high school
GPA does not consistently and sufficiently contribute to

predicting college success.

29

When an additional intellective criterion, normally
an aptitude test score, is added to the high school average,
the resulting multiple correlation with GPA is usually
higher than the correlation of either predictor alone.
Fishman and Pasanella (1960) reported multiple correlations
ranging from .37 to .83 with a median of .62. The average
gain in forecast over the high school grade point average
(as a single predictor) was found to be .11. When further
intellective measures were employed, however, only very
small gains in prediction were noted.

Webb (1967), also, found high school grades the best
single predictor of college grades, with the Scholastic
Aptitude Test verbal scores adding more than any other
variable to predictive efficiency. However, he reported
that personality variables contributed more than SAT scores
in predicting success in individual fields of study.

Dohner (1969) found that high school class ranks in com—

bination with American College Test (ACT) scores best pre-

 

dicted academic success.

Elton (1969) found the ACT mathematics score best in
predicting educational outcomes in females and the ACT
social studies score best in males.

Baird (1969) found that college grades were predicted
by self—ratings on scholarship and high school grades, with
the ACT social studies test improving the prediction in
males.

Studies into the use of both intellective predictors

and intellective criteria have been considered so important

30

that several articles have dealt with simplified methods of
predicting college grade point average (Aiken, 1968) and
college success (Merwin, 1964) from input variables such as

high school grade point average and the Scholastic Aptitude

 

Iggp. The Aiken article is of particular interest in that
it provides a graphic determination of, first, a triple
regression equation prediction of GPA and, next, an approx—
imation of the standard error of estimate, and, lastly, an
appropriate cutoff decision strategy.

The American College Testing Program (1965) reported
correlations between student-reported grades and corres—
ponding school-reported grades that ranged from .91 for a
large sample of ACT examinees. Although the predictive
validities of the two sets of grades were not directly com-
pared, a comparison of the predictive power of grades
reported by students and high school class ranks revealed
no consistent advantage for either variable.

Bogue (1963) found that student-reported grades used
with ACT scores of 372 examinees predicted college grades
slightly better than did school-reported grades with ACT
scores. Comparative predictive validity studies which used
examinees younger than high school seniors are not
available.

A particularly interesting aspect of predictability
is found in the differences of college grades in terms of
male and female students. Abelson (1952) and Seashore

(1962) have reported that a woman's GPA is significantly

31

more predictable (by intellective predictors) than is a
man's. A similar study by Paraskevopoulos and Robinson
(1970) indicated no significant differences in prediction
between sexes, but, instead, there were clear indications
that a separate regression equation for women was higher
(different Y-intercept) than was the mixed-sex regression
line. Thus, the combined prediction equation tended to
favor the male applicant since each female applicant's
predicted GPA was approximately .20 lower on the mixed-sex

regression equation.

B. Studies UsingyAffective Variables

 

As indicated earlier, multiple regression equations
with three or more intellective predictors did little to
increase the multiple correlation with GPA beyond .60.

But just as it appeared that research into intellective
factors had reached the point of diminishing returns in
work on prediction, a resurgence seems to have been stimu-
lated by the idea that some nonintellective measure might
provide a further explanation of the total variance of the
college GPA. In summarizing the studies of non-cognitive
variables in relation to academic achievement, Graff and
Hansen (1970) made the following statements:
A thorough review of the literature indicated that many
studies of the non-cognitive aspects of achievement
have been conducted during the last two decades.
Researchers tried to relate social background factors,
interests, Rorschach and TAT responses, study habits,
and different personality traits to academic achieve-

ment. Unfortunately, the results were generally
inconsistent or non-significant. Some of the

32

investigations produced correlations similar to those
found with conventional predictors of academic success.
The crucial issue, however, comes in determining how
much these nonintellectual components actually added

to the prediction validity based on high school records
and intellective tests. (P. 120.)

Research in recent years has tested a number of non-
intellective predictors of college success. An early study
by Hoyt and Norman (1954), for instance, indicated that an

"adjusted" student, as determined by his Minnesota Multi-

 

phasic Personality Inventory (MMPI) score, was significantly

 

more predictable than his "maladjusted" counterpart. A
later study by Anderson and Spencer (1963), however,
attempted to replicate the Hoyt and Norman results, but
found instead remarkable similarities in aptitude, achieve-
ment and predictability between the "adjusted" and "mal-
adjusted" groups.

Similar contradictory results were obtained in
various investigations into the nonintellective measure of
study habits and attitude questionnaires. Whitla (1969)
cites the Brown and Holtzman study which indicated that
their survey of study habits increased the multiple corre-
lation to about .70 when ability measures were also
included as predictors. As Whitla (1969) mentions, however,
a second study by Ahmann used the same inventory with no
significant increase in the predictive capacity of the
regression equation.

Recently, more encouraging findings have been reported
for studies which have approached the prediction of academic

attainment by means of prediction-oriented inventories

33

consisting of items which have been empirically selected
and empirically keyed. Illustrative of published scales

of this kind is the California Study Methods Survey by

 

Carter (1960), the Brown—Holtzman Survey of Studngabits

 

and Attitudes by Brown and Holtzman (1956), the California

 

 

Psychological Inventory by Gough (1957), and the Opinion

 

and Attitude Survey by Fricke (1960). Other scales,

 

exemplified by the work of Ward (1959) at the University

of Tennessee and Hebenstret (1959) at the University of
Washington, have been developed from a conglomerate of
items which assess biographical characteristics along with
study skills, personality, and motivation. Juola (1963)
made the criticism that some of the tests used, be they
measures of adjustment, interest, attitudes, values and so
forth, have been developed for purposes other than the
evaluation of academic adjustment and that it is therefore
not surprising that these non-cognitive inventories proved
totally inadequate as predictors of academic success. Then
he attempted to construct an empirically derived non-
cognitive scale that is based upon attitudes and values
that students hold for education and educational activities.
He reported that the correlation of his trial form of the

Academic Attitude Preference Inventory (AAPI) with first

 

quarter GPA for new freshmen was .52 for each sex and that
the correlation with the cumulative one-year GPA was .48.
Because "one of the basic assumptions in education is

that motivation is a prime requisite for scholastic success"

34

(McBee and Duke, 1960, p. 3), motivation has been exten—
sively studied (Heckhausen, 1967; McClelland, Atkinson,
Clark and Lowell, 1953). In his review of significant
research on the prediction of academic success, Travers
(1949) stated that motivational factors played a major role
in determining success both in high school and in college
and measures of interest had been found to correlate with
college performance almost as well as measures of aptitude.
Not only general motivation but also each of the specific
motivational variables was found to be significantly related
to academic performance. Atkinson (1958) cautioned that it
is unlikely that academic achievement may be predicted by a
measure of a single motive since it may satisfy more than
one need, such as understanding, power, or affiliation.
Furthermore, McKeachie, Isaacson, Milholland and Lin (1968)
observed that most of the successful studies relating
achievement motive to academic achievement have been done
with males, and Klinger (1966) observed more frequent occur—
rence of significant relationships between need for achieve—
ment and academic performance among secondary school stu-
dents than among college students.

Holland (1966) suggested the usefulness of brief
lists of activities and brief lists of competencies for
predictors. Baird (1969) found that trait self-ratings and
self—evaluation of life goals might be useful predictors.
However, it is very difficult to study most of the psycho—

logical constructs, because they are intervening variables

35

and not directly measurable. Lack of agreement on the
definition of the construct adds difficulty in designing
the study and interpreting the results of the study.
Correlations of interests with grades in related
fields are generally below .30, so interest tests add only
a small amount to academic prediction (Cronbach, 1970).
Interests sometimes predict who stays in training and who
drops out. Of those with A and B+ scores on the Dentist

key of the Strong Vocational Interest Blank, 92 percent

 

graduated from dentist training, compared with 67 percent
of B's and 25 percent of C's (Strong, 1943, p. 524). It
has been suggested that the profile will predict differ-
ences in grades between preferred and non-preferred areas.
An extensive study by French (1958) shows that prediction
of this sort has too little accuracy to be of use.

There have been other attempts to employ nonintellec-
tive characteristics. Berdie (1961) found significantly
lower predictability for those students whose intra—
individual variability on a preadmission test was higher
than for those whose variability was low.

Barclay (1965) described Bendig's study which was

carried out using the Edwards Personal Preference Schedule

 

(EPPS) as a predictor in a multiple regression equation.
The use of the EPPS added .09 to the multiple correlation
in this case. Frederiksen and Melville (1954) indicated

that a student rated "compulsive" on Strong Vocational

 

Interest Blank is significantly less predictable on his GPA

 

than is the "noncompulsive" student.

36

In most cases, however, the relationships between
personality variables and academic criteria have been found
to be quite low. An important recent exception is the
Holland and Nichols (1964) study which took a group of
students highly homogeneous with respect to intellectual
ability (all National Merit Scholarship finalists) and
examined them in terms of nonintellective predictors.

These nonintellective characteristics were found to be
effective predictors of college grades. Two important
groups of prediction-related personality factors were iden-
tified: (a) motivation to succeed and (b) conformity to a
basic socialization and value system. The significance of
this experiment is that nonintellective measures were

shown to have significant predictive validity when academic
ability is held relatively constant.

Finally, much progress has been made in order that
students can be helped to improve their academic perfor—
mance. Baymur and Patterson (1960) and Hatch, Dressel and
Costar (1963) pointed out that students benefit from appro-
priate counseling. Working with urban school, ninth—grade
underachievers with low self—concept, Brookover (1962, 1965)
found that they could be helped to achieve significantly
better by improving their self-concepts through the efforts
of "significant others." Wrenn and Humber (1941) suggested
that improving students habits may help them improve scho—
lastically, and Stebens (1957) believes that programs for

reading skill improvement are very fruitful.

37

C. Studies Using Other Variables

 

Biographical inventories were used in an attempt to
find nonintellective factors which would significantly add
to prediction of criterion variance above and beyond that
accounted for by intellective test measures (Cosand, 1953).

In 1911, Pittenger surveyed the freshmen grades of
the within-state students at the University of Minnesota.
His conclusions were that the graduates of the larger high
schools might be expected to do better than those from the
small school.

At the State College of Washington, Thornburg (1924)
studied freshmen grades on the same variable. He concluded
that students from the large high schools achieve superior
grades at college. But, in 1949 at Purdue, White (1951)
made a study of the University's admission criteria. His
findings revealed that size of high school had no relation-
ship to college success at that state university. Shafer
(1956), in a study of students entering certain Iowa
colleges, came to the same conclusions as did White. This
investigation was carried out in 1956. The size groupings
used were: 0—99; 100-199; 200—299; 300-499; 500-999; and
1,000 and above.

In 1962, Harmon reported on the relationship between
doctorate productivity and size of high school graduating
class. The author pointed out that apparently something
was happening in the high schools to differentiate the

people who, more than a decade later, will earn doctoral

38

degrees in various scholarly fields. He concluded that
size of high school, as reflected in size of graduating
class, has a profound effect on the probability of an indi-
vidual's going on to college, to graduate school and even-
tually to the doctoral degree.

Bloom (1964) referred to the six variables in the
home environment: "achievement press," language models,
academic guidance, stimulation to explore various aspects
of the larger environment, intellectual interests and
activities, and work habits emphasized. Dave (1963)
obtained a correlation of .80 between ratings of home
environment on these variables and achievement test battery
scores of children. The usual correlations between the
socio-economic status and achievement are less than .50.
Kurtz and Swenson (1951) found that home environment had
some influence on achievement. Where parents show interest
and pride in their children and children wish to please
parents, there seems to be more achievement.

Smith (1965) studied 154 University of Kentucky male
freshmen to determine differences between high-ability
achieving and non—achieving students. Students in his
sample scored in the upper fifth percentile on the College

Qualification Tests. He tentatively concluded that stu-

 

dents who came from larger metropolitan areas possessed a
set of values and attitudes concerning education which

seemed to make them more prone to underachievement.

39

Staton (1962) studied new freshmen from Oklahoma high
schools who enrolled in the University of Oklahoma. One of
the variables he selected for study was student's high
school curriculum. He concluded that the curriculum taken
in high school did not influence college grades. Young
(1967) analyzed the high school curriculum patterns of
closely matched pairs of college students. He found no
significant difference in college achievement between stu-
dents who took 7.9 business and industrial courses in high
school and students who took 0.9 such courses.

D. Studies Related to Some Other
Problems of Prediction

 

 

The basic assumption of the regression formula
employed in the statistical analysis of the prediction data
is that there is a linear relationship between the predic-
tors and the criteria. Weiss's recent paper (1970) argues
that the relationship is more likely non-linear. He
employed a non-linear assignment of weights and obtained a
Spearman's rank order correlation coefficient between
weighted predictors (intellective) and GPA of .89. Although
Weiss admits this is but a "first approach to developing a

non-linear predictive system,‘ it does appear that the non-
linear approach demands some concentrated investigation.

Recent emphasis has been on the use of a combination
of variables. There is a distinct superiority in multi—

variable prediction over prediction by the use of a single

factor. Cosand (1953) summarized studies of multiple

4O

correlations. These correlations point out the advantage
of using several predictors rather than a single one.

Spiegel (1971) sought to use a stepwise multiple
regression method to select from intellectual, attitude,
and personality variables linear combinations of variables
that might optimize prediction of course points in male
and female first year college students. Fifty—four vari-
ables were used in the analysis. For males, twelve vari-
ables were included in the predictor set which yielded a
coefficient of .85. For females, ten variables were
included in the predictor set that yielded a coefficient
R of .92.

The predictor measures studied usually serve as
admissions criteria. Also, students make decisions to
apply to one college or to another college. These facts
might affect the predictability of GPA because they are
closely related to the range of the population in a college.
It is interesting to note here that Fishman and Pasanella
(1960) found that the highest reported correlations (high
.60's and .70's) were all obtained from Southwestern and
Western colleges in which selection procedures were minimal.

Low predictability of college success might be due to
the fact that the criterion, grade point average, is a very
complex and not very valid and reliable measure. Whitla
(1969) mentioned that a freshman might have a choice of
some 148 possible courses in 39 departments at Yale. In

light of this possible variability, to equate one

41

freshman's GPA with another's is tenuous at best. One
other source of the unreliability of grades arises from
the variability in grading systems that are prevalent in
schools. A student with an "A" grade in one college may
be only as able as, or perhaps less able than, a student
with a grade of "B" in another college.

Various techniques have been employed to correct for
this variability while predicting college achievement from
school grades, as discussed by Bloom and Peters (1961).
Linn (1966) reviewed the results of several empirical
studies that have used "adjusted" grades to predict aca—
demic achievement. His paper considered some of the pos-
sible techniques which could be used to make grade adjust-l
ments for interschool differences. Most researchers,
however, have found that the improvement in predictive
validity due to the use of adjusted grades has been dis-

couragingly small.

Summary of the Chapter

 

This chapter consisted of a critical examination of
the literature concerned with the construct of intellectual
interest, college as a socializing institution and predic-
tion of college success.

With several exceptions, there were few attempts to
investigate the construct of intellectual interest. Some
researchers have developed scales to measure intellectual
interest and have attempted to reveal the characteristics

of the domain. Yet the clear identity and structure of

42

intellectual interest which is the main domain of the
present study is not well provided.

In spite of the vagueness of the concept, the domain
has been widely employed in studies attempting to under-
stand students, college environment and the role of higher
education. Moreover, some attempts to validate the domain
of intellectual interest against college success have been
made.

The literature contains studies of academic perfor-

mance at all educational levels and that which pertains to
undergraduates in colleges is particularly voluminous.
The use of tests in the selection of applicants for admis-
sion and in the prediction of academic success, defined in
terms of college grades, has been the most explored topic
in educational and psychological research.

Many studies indicate that high school grade point
average is the best single predictor of college success.
With an additional intellective predictor, normally an
aptitude test score, the resulting multiple correlation
with GPA has usually been significantly higher than the
correlation of either predictor alone.

Many researchers have also investigated personality,
biographical and demographic variables, primarily in an
attempt to increase the predictive efficiency of students'
college achievement. The results of many studies have

often been inconsistent and, sometimes, contradictory.

43

Many studies have attempted to isolate non-cognitive
correlates of college success. While a considerable num-
ber of non-cognitive variables have at one time or other
been correlated with student achievement, the direction
and magnitude of the relationships have generally not been
consistent from study to study.

Despite the many studies which have been done con-
cerning the global prediction of grades, little progress
has been made in the prediction of college success by means
of multiple regression techniques, differential prediction,
moderated regression models and non-linear relationships.

The method of this study, along with the statistical
hypotheses to be tested, will be found in Chapter III

which follows.

CHAPTER III

METHODOLOGY

Population and Sample

 

The population examined in this study consisted of
all freshmen who entered Michigan State University Fall
term, 1970. Five thousand four hundred and sixty-eight
freshmen students registered for credit courses during the
fall registration period. However, some restrictions were
imposed on the population. The following types of students
were excluded:

1. All students who previously attended any college

or university,

2. All foreign students,

3. Students whose test data were incomplete, and

4. Students who dropped out before the end of Fall

term, 1970.

The sample of students used in the study was selected
from the restricted population as defined above, and any
future references to the population of the study, or
generalizations and conclusions to be drawn from the
results of the analysis should be interpreted in terms of

the restricted pOpulation.

44

45

Out of the restricted population, 643 students were
randomly selected for study. Because of this random selec—
tion procedure, the sample of the above 643 students could

represent the total restricted population of the study.

Instrumentation and Criterion

 

The study employed several instruments related to
cognitive, affective, demographic, and background charac-
teristics. In addition to them, the composite scale was
made to measure the construct of intellectual interest.

A list of the instruments used in the study follows:

A. Instruments used to measure the construct of

intellectual interest:

1. Four subscales from the Stern Activities

 

Index, including Reflectiveness, Humanities-
Social Science, Understanding, and Science;

2. Intellectual Interest Scale of Anderson and

 

Western; and

 

3. Intellectualism—Pragmatism Scale of Yuker and
Block.

B. Instruments used to measure the affective charac-
teristics:

1. General Self-Concept of Academic Ability;

 

2. Trait Self—Ratings of College Freshmen; and

 

3. Life Goals of College Freshmen.

 

C. Instruments used to measure the cognitive charac-

teristics:

46

l. Scholastic Aptitude Test;

 

2. Michigan State University Reading Test;

 

3. Michigan State University Arithmetic Test

and Mathematics Test.

 

D. Instruments used to measure demographic and back-
ground characteristics:

1. Student Questionnaire.

 

The somewhat detailed information for each of these

instruments is discussed in the following pages.

A. Intellectual Interest Scale

 

Four subscales from the Stern Activities Index,

 

namely, Reflectiveness, Humanities-Social Science, Under-

standing, and Science; the Intellectual Interest Scale

 

developed by Anderson and Western; and the Intellectualism-

 

Pragmatism Scale developed by Yuker and Block were employed

 

to measure the psychological construct of intellectual
interest. The composite scale formed from all of the

above three scales was named the Academic Interest Scale

 

in order to deter faking responses. Descriptions of each
of the three scales are given below.

1. Intellectual Interests Scale of Stern Activities

 

 

Index.

In developing the Activities Index, Stern (1963a)

 

viewed the college as a system composed of a number of
interdependent parts which share, to one degree or another,
certain values and characteristics. Furthermore, the

college is viewed as a social system in the sense that the

47

parts involve people-~there are individual and group needs

to be satisfied. The Scale consists of three hundred

items and each subscale has ten items. Stern (1963a)

defined intellectual interest which is one aspect of factor

three, Intellectual Orientation Dimension, as follows:
Factor 3. Intellectual Interests. The factors with

the highest loadings in this dimension are based on
items involving various forms of intellectual activi-

ties. These include interests in the arts as well as
the sciences, both abstract and empirical. Score sum:
Reflectiveness, Humanities—Social Sciences, Under—
standing, and Science. (P. 14.)

Extensive data for reliability and validity are pro—
vided. For the purpose of the present study, four sub-
scales, i.e., Reflectiveness, Humanities-Social Sciences,
Understanding and Science, were chosen to make the Intel-

lectual Interest Scale. This scale covers items numbered

 

one through forty on pages 1 and 2 of the Academic Interest

 

Scale in Appendix A. The items are distributed in the

Academic Interest Scale as follows: items 1 to 5 and 21 to

 

25 cover the subscale of Reflectiveness; 6 to 10 and 26 to
30, Understanding; 11 to 15 and 31 to 35, Science; and 16
to 20 and 36 to 40, Humanities—Social Sciences.

The scoring system of the scale was that a score of
one was assigned to the response "Blacken space l——if the
item describes an activity or event that you would 11kg,

enjoy, or find more pleasant." A score of zero was given

 

to the items having the response "Blacken space 2--if the
item describes an activity or event that you would dislike,

reject, or find more unpleasant than pleasant." The Scale

 

score was computed by summing raw item scores.

48

2. Intellectual Interest Scale of Anderson and

 

Western.

Anderson and Western (1966) developed a definition
of intellectual interests as a dimension of appreciation
and enjoyment of cultural pursuits and an interest in
philoSOphical discussion and discourse. The emphasis of
the dimension is based on a liking for, but not necessarily
sustained activity in, certain pursuits-—hence intellectual
interests.

The Scale consists of nine items answered by choosing
a category which indicated degree of agreement, or the
extent to which a statement is true, resulting in a four-
point response key.

Concerning all of the subscales of An Inventory of

 

Students' Attitudes, including the Intellectual Interests

 

 

Sgale, the authors provided the inter-scale correlations

found in Table 3.1.

TABLE 3.1.--Inter-Scale Correlations of An Inventory of
Students' Attitudes by Anderson and Western

 

 

 

 

Scale 2 3 4 5 6 7
1 (Intellectual Interest) -.23 .36 —.49 .15 .15 -.17
2 (Dogmatism) —.33 .24 -.25 —.05 —.19
3 (Tolerance of Complexity) -.37 .19 .12 .15
4 (Pragmatism) -.17 -.ll -.13
5 (Social Liberalism) .02 .34
6 (Economic Liberalism) .27
7 (Political Liberalism)

 

The Intellectual Interests Scale consists of items

 

numbered 41 through 49 on page 3 of the Academic Interest

 

Scale in Appendix A.

49

This scale is a nine-item Likert-type scale. "Defi—
nitely true" responses were given a weight of 4, "More true
than false" was given a weight of 3, "More false than true"
was given a weight of 2, and "Definitely untrue" a weight
of l. A scale score is the sum of each of the item scores.

3. Intellectualism-Pragmatism Scale (I—P Scale) of

Yuker and Block.
While the instrument, developed by H. E. Yuker and

J. R. Block, was originally referred to as The Attitude

 

Toward Intellectualism Scale, the authors renamed it the

 

Intellectualism-Pragmatism Scale (I-P Scale).

 

They stated their motivation for developing the Scale
as follows:

Although intellectualism is often discussed, there have
been comparatively few attempts to develop a measure of
intellectual attitudes. Most of us are apparently con-
tent to discuss and speculate about these attitudes
without operationally defining them, or attempting to
empirically determine any of the correlates of intel—
lectualism. The present attitude scale was developed

in order to provide an empirical measure of intellectual
attitude. (Yuker and Block, 1969, p. 1)

Inspection of items of the Scale, according to the
authors, indicates that they all have face validity as
measures of intellectual-pragmatic attitudes.

Reliability coefficients, as estimated through the
split—half technique corrected using the Spearman—Brown
formula with different samples of undergraduate college
students, tend consistently toward the mid—eighties with a

median of approximately .84. The only evidence of test-

retest reliability is available for a group of thirty

50

undergraduates enrolled in a course in introductory psycho-
logy at Hofstra University. The coefficient of reliability
when an interval of four months elapsed between test admin—
istrations was .84.

Construct validity was used in evaluating the adequacy
of the I-P Scale. Only education-related variables have
been reported. A correlation coefficient of .56 was found
between I-P scores and grade point average for a sample of
120 evening students. Low, but significant, positive
correlations were found between I-P scores and scores on
the verbal part of the SAT, and between I-P scores and
scores on a measure of reading ability.

The I—P Scale consists of items numbered 93 through

121 on page 4 of the Academic Interest Scale in Appendix A.

 

This scale, a thirty—item Likert-type attitude scale,
was developed to measure a continuum of intelectual versus
pragmatic attitudes. Some of the statements were worded so
that agreement would indicate an intellectual attitude
while others were worded so that intellectualism would be
reflected by disagreement. This latter type of item
included itsms 96, 97, 98, 100, 101, 102, 103, 107, 111,
112, 118, 119 and 122.

Responses contained six categories of agreement and
disagreement ranging from +3 to -3. Scoring of the test
was accomplished by changing the algebraic sign of the
subject's responses to the above fifteen items which are

negatively worded. To eliminate negative numbers, a linear

51

transformation was made by adding 3 to each item. In other
words, "Very strongly agree" responses were assigned a
weight of 6, "Strongly agree" a weight of 5, "Agree" a
weight of 4, "Disagree" a weight of 3, "Strongly disagree"
a weight of 2 and "Very strongly disagree" a weight of 1.

The response categories and scoring system of each
of the three subscales used to measure the construct of
intellectual interest were described earlier. Scores on
three separate and independent subscales were obtained to
provide a cross-check on their validity. If only one
method of measurement is used as a basis for estimating the
strength of a trait, there is no check on the validity of
the measure relating trait to behavior. When more than one
measurement of a trait is used, confidence in the construct
and in the methods for measuring it increases when the
intercorrelations among the several sets of scores are high.
Such results would suggest the various methods of measure-
ment converge on a simple trait.

These are some problems involved in deriving a com-
posite score for the three scales together. This is
because the three subscales do not use the same response
categories and scoring systems. Since the three subscales
have different response categories and scoring systems,
several efforts were made to derive the most reasonable
composite score. The eleven variables listed below were
generated from these efforts. The variables, which are

also used in Table 3.2, were named on the basis of the

52

different subscales and scoring systems used to derive the

composite score.

Variable 1: Stern-Reflectiveness based on 2-point
scoring system.

Variable 2: Stern-Understanding based on 2-point
scoring system.

Variable 3: Stern-Science based on 2-point scoring
system.

Variable 4: Stern-Humanities-Social Sciences based
on 2-point scoring system.

Variable 5: Stern-Total score based on 2-point
scoring system.

Variable 6: Anderson and Western-Total score based
on 4-point scoring system.

Variable 7: Yuker and Block-Total score based on
6-point scoring system.

Variable 8: AIS-Total score consisted of summated
score of variables 5, 6, and 7.

Variable 9: ~Anderson and Western-Total score based
on 2-point scoring system.

Variable 10: Yuker and Block-Total score based on
2-point scoring system.

Variable ll: AIS-Total score consisted of summated
score of variables 5, 9, and 10.

TABLE 3.2.--Intercorrelation Coefficients of Variables

Based on the Different Subscales and Different
Scoring System of the Intellectual Interests

 

 

 

Scale
Variable (l) (2) (3) (4) (5) (6) (7) (8) (9) (10)
1
2 .32
3 .20 .52
4 .31 .16 .03
5 .62 .76 .72 .55
6 .43 .28 .08 .56 .48
7 .30 .23 .09 .29 .33 .46
8 .50 .46 .30 .50 .65 .69 .91
9 .40 .29 .06 .54 .46 .93 .44 .66
10 .32 .26 .10 .30 .35 .50 .86 .83 .49
ll .61 .66 .52 .59 .89 .70 .64 .88 .70 .74

 

The correlation coefficients in Table 3.2 led

to the development of a rationale for the derivation of a

53

composite score and also provided the evidence of the con-
vergent validity of the subscales.

Table 3.2 shows that the correlation coefficient of
the composite score (variable 11) was .89 with the Stern-
Total score, .70 with the Anderson and Western-Total score
and .64 with the Yuker and Block-Total score. Even though
these scores represented independent efforts to measure the
construct of intellectual interest, they were found to
agree with one another, as shown by the high inter-correla—
tions among them. It was reasonably certain that they were
assessing the same trait with accuracy.

The higher correlation value of variable 11 with the
rest of the variables justified the use of a 2—point
scaling system for all of the subscales. Furthermore, all
of these three subscales were summated rating scales.
Items in a summated rating scale or Likert—type scale are
considered to be of approximately equal value. Any subset
of the universe of items is theoretically the same as any
other subset of the universe. This problem is related to
the response variance. Because response variance was
determined by the number of possible categories, response

set should be the same for all items.

B. Michigan State University Student Survey

 

1. General Self—Concept of Academic Ability.

 

This scale was developed by W. B. Brookover (1965)
for his study of self-concept of ability and achievement

of senior high school students. It consists of eight

54

five-choice items. Items were coded from five to one with
the higher self-concept alternatives receiving the higher
values. These eight items were originally written to form
a Guttman scale and received coefficients of stability of
.95 for males and .96 for females in the tenth grade. High
correlations of self-concept of ability in subject matter
areas with general self-concept of ability supports the
validity of the instrument. Although no direct data were
provided for either twelfth graders or college freshmen,
some of the author's unpublished data suggest that the
scale would be valid for use with college freshmen. The
scale is shOWn as Part C (items numbered 43 through 50) of

the M.S.U. Student Survey in Appendix B.

 

2. Trait Self-Ratings of College Freshmen.

 

This factor analytic scale, developed by J. M.
Richards, Jr. (1966), grew out of the American College
Survey which was conducted by the American College Testing
Program to obtain a more complete description of the typi-
cal American college student and the variation among stu—
dents from college to college. This scale consists of
thirty—one self—ratings on common traits for both sexes.
Each of the subjects rates himself or herself on each of

the thirty-one traits using the four—point response key:

"Below Average," "Average," ”Above Average," and "Top Ten

Percent." Scores from 1 to 4 were assigned to these

responses so that a higher score indicates a greater pos-

session of the trait in question. Seven factors were found.

 

55

In adapting the scale for this study the three items
having the highest loading for each factor were selected.
The same number of items made comparing each factor much
easier.

The scale has twenty-one items. Correlations among
Promax Oblique Factors provided by the author of the scale
are found in Table 3.3.

TABLE 3.3.-—Correlations among Promax Oblique Factors of
Trait Self—Ratings of College Freshmen*

 

 

 

Factor** A B C D E F G
A .24 .26 .39 —.ll .51 .03
B .27 .39 .46 —.12 .26 —.09
C .44 .21 .52 -.24 .45 -.04
D .31 .39 .30 -.25 .46 -.13
E .46 .33 .22 .33 -.23 .19
F -.03 .09 .08 ~.07 -.01 -.02
G .38 .19 .32 .29 .25 -.Ol

 

*Correlations for males are shown above the diagonal
and for females below. Factors are reflected as appropriate.

**A-—Physica1 well-being; B-—Scholarship; C--Estheti-

cism; D-—Pragmatism; E-~Technical—scientific ability;
F-—Sociability; and G--Sensitivity to others.

The Trait Self—Ratings of College Freshmen scale is

 

shown as Part A (items numbered 1 through 21) of the M.S.U.

Student Survey in Appendix B.

 

3. Life Goals of College Freshmen.

 

J. M. Richards, Jr. (1966), using a sample of 6,289
male and 6,143 female college freshmen, developed 35
items pertaining to life goals of college freshmen. This

scale, like Trait Self-Ratings of College Freshmen, grew

 

56

out of the American College Survey project conducted by the
American College Testing Program in an attempt to obtain a
more complete description of the typical American college
student. Each of the thirty-five specific life goal items
is rated by the subject on a four-point scale such as

"Of little or no importance," "Somewhat important," "Very

important,‘ and "Essential for me." Scores from 1 to 4
were assigned to the responses so that a higher score indi-
cates a greater possession of the life goal in question.
Seven factors common to both sexes and one unique factor
for each sex were found. Common factors are prestige,
personal happiness, humanistic-cultural, religious, scien-
tific, artistic and hedonistic. The unique factor for male
is athletic and that for female is altruistic. The present
edition of the scale includes only the seven factors common
to both sexes. For each factor, the three items having the
highest factor loading on it were selected.
Intercorrelations of the factors for the sample of
the present study are given in the following table (3.4).
The data show that these seven factors are substantially

independent.

The Life Goals of College Freshmen scale is shown as

 

Part B (items numbered 22 through 42) of the M.S.U. Student

 

Survey in Appendix B.

4. Student Questionnaire.

 

The original form of the Student Questionnaire was

 

developed by the Office of Evaluation Services, Michigan

57

TABLE 3.4.—-Intercorrelations of the Factors of the Life
Goals of College Freshmen for the Present

 

 

 

 

Sample
Factor* 1 2 3 4 5 6 7
l
2 .07
3 .29 .14
4 .22 .25 .20
5 .16 -.10 .10 .08
6 .10 .14 .17 .14 .09
7 .30 .21 -.02 .01 .06 .13

 

*l--prestige; 2--personal happiness; 3--humanistic—
cultural; 4--religious; 5~-scientific; 6--artistic; and
7--hedonistic.

State University. It consisted of sixty items relating to
biographical, demographic and background information, and
opinions related to current social issues. For the purpose
of the present study, several relevant items were chosen.
This short—form questionnaire was used to obtain the bio-
graphical, demographic and background information about
each student in the sample.

The Student Questionnaire is shown as Part D (items

 

numbered 101 through 132) of the M.S.U. Student Survey in

 

Appendix B.

C. Scholastic Aptitude Test

 

The Scholastic Aptitude Test (SAT) of the College

 

Entrance Examination Board assessed the basic verbal and
mathematical abilities a student has acquired. This test,
which is generally administered to college-bound students

in the senior year of high school, assesses the ability

58

to reason rather than to remember facts and requires no
special preparation. The student receives three scores,
Verbal (SAT-V), Mathematical (SAT-M), and Total (SAT-Total).
These scores are reported nationally as three-digit stan-
dard scores where 500 was initially set as a national
college-bound senior average score. Each score can be as
high as 800, thereby making 1,600 the maximum possible

total score. Test—retest reliability coefficients of .89
for the Verbal scale and .85 for the Mathematical scale

are reported. Much validity data have been published.

D. Michigan State University Reading Test

 

The MSU Reading Test was developed by the Office of

 

Evaluation Services, Michigan State University. The Test
was designed to measure a student's ability to comprehend
ideas expressed in paragraphs representative of those found
in textual materials of various academic areas at MSU. The
Test consists of 50 items and is used on a supplementary
basis for selecting students for the Preparatory English
Program as well as for selection into honors programs.
Reliability of the Test has been estimated on several
occasions by the Office of Evaluation Services to be approx—
imately .80. Correlation between the Reading Test and
first quarter GPA for the sample in this study was .45 for

males and .49 for females.

59

E. The Michigan State University Arithmetic
Test and Mathematics Test

 

 

The MSU Arithmetic Test is a forty—item test of

 

elementary arithmetic problems. Scores range from zero to
forty. Students are assigned to an "Arithmetic Improvement
Service" course when their scores are 24 or lower.

The MSU Mathematics Test is a thirty-item test based
upon concepts covered in high school algebra. The scores
are used to place students in beginning courses in mathe-
matics. Scores range from zero to thirty.

The composite score which was called the MSU Quanti-
tative score is based upon the seventy-item sum of the
Arithmetic and Mathematics scores. This measure is indica-
tive of directly applied quantitative skills and knowledge.
It differs from the SAT Mathematics score which emphasizes

the power to reason with quantitative concepts.

Criterion

 

The criterion employed in this study was the cumula-
tive grade point average at the end of the Fall term, 1971.
This criterion was assumed to be the best indication of the
student's academic standing after completing one term at
the University. The Michigan State University grade system
is a lO-point scale ranging from zero to 4.5. Each stu-
dent's numeric grade is multiplied by the number of credit
hours in the course for which the grade was given. The sum
of the products of numeric grade by credit hours for all of
the student's courses is divided by the sum of the credit

hours. The quotient is the student's grade point average.

60

Collection of Data

 

During the orientation week, September 21-22, 1970,
the following instruments were administered to the sample
of students in the study at Michigan State University:

the Academic Interest Scale and the MSU Student Survey.

 

 

As pointed out in the "Instrumentation" section above,

the Academic Interest Scale contains the Intellectual

 

 

Interests Scale of Stern (Part 1); the Intellectual Inter-

 

 

est Scale of Anderson and Western (Part 2); and the

 

Intellectualism-Pragmatism Scale of Yuker and Block (Part

 

3). The MSU Student Survey contains several scales, includ-

 

ing the Trait Self—Ratipgs of College Freshmen (Part A);

 

the Life Goals of College Freshmen (Part B); the General

 

Self-Concept of Academic Ability (Part C); and the Student

 

Questionnaire (Part D).

 

The results of Scholastic Aptitude Test, the MSU

 

Reading Test, the MSU Arithmetic Test, the MSU Mathematics

 

 

 

Test, the MSU Quantitative Test and high school grade point

 

average were obtained in December of 1970 with the cooper—
ation of the Office of Evaluation Services, Michigan State
University. The college cumulative grade point averages
of the subjects which form the criterion of the measure-
ment in the validity part of the study were also made
available to the writer through the cooperation of the

Office of Evaluation Services.

61

The Statistical Modelg

 

After a review of several possible statistical models
the factor analytic method was selected as the most appro-
priate technique for the purpose of reaching a clear under-
standing of the structure and pattern of intellectual
interest. Factor analysis is a means by which the regular-
ity and order in phenomena can.be discerned. It can be
applied in order to explore a content area, structure a
domain, map unknown concepts, classify or reduce data, test
hypotheses, formulate theories, or make inferences. Factor
analysis is most familiar to researchers as an exploratory
tool for discovering the basic empirical concepts in a
field of investigation. Representing patterns of relation-
ship between phenomena, these basic concepts may corrob-
orate the reality of prevailing concepts or may be so new
and strange as to defy immediate labeling.

Factor analysis is often used to discover such con-
cepts reflecting unsuspected influences at work in a domain.
The delineation of these interrelated phenomena enables
generalizations to be made and hypotheses to be posed about
the underlying influences bringing about the relationships.
The factor analytic technique was supplemented by analysis
of variance for several qualitative variables. In the
analysis of variance, hypotheses were tested at d = .01.

The purpose of the second part of the study was to
measure the validity of intellectual interest in the pre-

diction of college success. Several procedures were

62

employed. Since there is no guarantee that all psycho-
logical relationships of either theoretical or applied
nature are linear in form, testing for linear and non-
linear regression was carried out to answer this question.
Also, the relationship between the criterion and the pre-
dictive variable was plotted on a scatter diagram and
graphically examined for possible departures from linearity.

After testing the regression model, the predictive
validity of the construct of intellectual interest in the
prediction of college success was assessed. The validity
of the intellectual interest scale was also determined. A
simple regression equation was also developed with intel—
lectual interest as the predictor and college grade point
average as the criterion variable.

The multiple regression technique was also used to
determine the ability to predict using several variables
simultaneously. For practical prediction situations in
college admissions, it is seldom the case that only one
item of prior information is known about the individual
subject. Usually several tests are given. The admission
officer in a college may have college entrance scores,
high school achievement test scores, high school grade
point average, and a great many other items of information
about individual students.

Under such circumstances, it is valuable to know
whether information regarding intellectual interest

improves the prediction of college success when it is

63

added to other predictors. For this purpose of determining
the incremental validity, the variance-ratio technique was
utilized. A detailed description of the statistical models

used in the study is presented below.

A. Factor Analytic Method

 

The principle axis method of factor analysis devel-
oped by Hotelling (1935) was used because it gives the
smallest number of factors which extract the maximum amount
of variance with a mathematically unique solution. Because
unities were used in the diagonal of the matrix, all vari-
ance, reliable or unreliable, was factored. The results of
the factoring method include all variance in the sample
with the unreliable variance randomly distributed among
factors. The factors do not refer to the population but
to the empirically functioning components within the sample.
The components describe the source of the variance. The
principle axis method gives a unique resolution of the
common factors or components for each sample when unities
are inserted in the diagonal. Factor loadings define
(l) a pattern of relationship and (2) the association of
each characteristic with each pattern. In general, for any
of the Y variables of equations, we may write:

= + +ooo+
Y le1 d2F2 dem ,

with the F's representing factors and the d's representing
loadings. These common factors aid in the interpretation
of the construct of intellectual interest because they are

based on empirical observation.

64

The further facilitation of an interpretation was
offered by rotation. The purpose of the rotation was to
transform the initial factor solution to a "preferred”
solution to achieve simple structure, factor invariance,
and interpretability. The unrotated factors successively
define the most general patterns of relationship in the
data. This is not so with the rotated factors. They
delineate the distinct clusters of relationships, if such
exist. Each solution is correct but psychologically cer-
tain solutions are preferred as being more interpretable.

Two objective methods of rotation were available:
the Quartimax method of Wrigley and Neuhaus (1954) and the
Varimax method of Kaiser (1959). Both methods attempt to
achieve simple structure principles based on the following
criteria of Thurstone(1947).

1. Each row of the factor matrix should have at
least one zero.

2. If there are m common factors, each column of the
factor matrix should have at least m zeroes.

3. For every pair of columns of the factor matrix,
there should be several whose entries vanish in
one column but not in the other.

4. For every pair of columns of the factor matrix,
a large proportion of the variables should have
vanishing entries in both columns when there are
four or more factors.

5. For every pair of columns of the factor matrix,
there should be only a small number of variables
with non-vanishing entries in both columns.

A simple structure rotation has several characteris-

tics that are of interest here:

65

1. Each variable is identified with one or a small
proportion of the factors.

2. The number of variables loading highly on a
factor is minimized.

3. A major ontological assumption underlying the

use of simple structure is that, whenever pos—
sible, our model of reality should be simplified.

4. A goal of research is to generalize factor

results. The unrotated factor solution, however,
depends on all the variables.

In comparison to the Quartimax method, which stresses
the simplification of each row or variable, the Varimax
method places emphasis on the simplification of factors.

To quote Harman (1960), "The Varimax method proposed by
Kaiser is a modification of the Quartimax method which more
nearly approximates simple structure" (p. 304). For this
reason, the principle axis and the Varimax methods were
employed to identify the structure and pattern of intellec-
tual interest.

B. Simple and Multiple Regression
Technique

 

 

When we want to predict the relative status of an
individual on the criterion variable a regression equation
provides the best estimate in terms of minimal squared
error.

When Y is the predicted score and X is the known
score on the independent variable, it is simple to develop
the regression equation for prediction of Y from X. The
linear model which was applied in the present study takes

the form of:

66

Yij = “Y + By-x (Xj - uX) + eij ,
where
= 91
By.x pxy ox

which is called the simple regression coefficient of Y on X.
This model has the following basic assumptions:
1. Within each population j, the distribution of
Yij values is normal;
2. Within each population j, the variance 0: is the
same; and

3. The errors eij are completely independent.

Given a random sample, the value of the sample regres-
sion coefficient byx is our best available estimate of

B

y x' the population regression coefficient. Moreover, the

best estimate of By- (ux) is given by b (MX). Our best

x yx

estimate of My is simply My.

Furthermore, it is important to be able to predict
the value of Y given the combination of several variables
considered simultaneously.

In general, in K—variable problems, the squared value
of the multiple correlation coefficient turns out to be:

2

R 1.2---K= (b r

r + ... + (biK. ... K—1)

12 1K '

12.3---K)
where R1.2 --- K denotes the correlation between a weighted
combination of independent variables and the criterion vari-
able. The squared multiple correlation coefficient indi~
cates the proportion of variance in the criterion variable

Y accounted for by the set of K predictor variables. Also,

67

the hypothesis about the multiple regression equation was
tested to see if the addition of intellectual interest

score really increased the value of R2.

C. The Variance—ratio Test

 

The variance-ratio test was used to test the incre—
ment in the criterion variance when the intellectual
interest score was added to the other predictor variables.
The test for significance was suggested by Baggaley (1962).
The ratio is given by:

2 2
F = (R+ — R ) (N - m - 2)

2
1 - R+

 

where R is the multiple correlation involving m predictors

and R+ is the multiple correlation involving m + 1 pre-

dictors. The quotient should be referred to an F table
with d.f. = l for the "greater mean square" and d.f. =
N - m - 2 for the "lesser mean square."

Statistical Hypptheses

 

In order to make the statistical tests of signifi—
cance, the following testable null hypotheses were formu-
lated from the previously stated purposes of the study and
substantive hypotheses.

Hypothesis 1: There is no difference in intellectual
interest, as measured by the Academic

Interest Scale, between males and
females and among:

 

 

a. Students majoring in different curricula;

b. Students who lived a major portion of
their lives on a farm, in a village,
town, small city, or large city;

Hyppthesis 2:

 

Hypothesis 3:

 

Hyppthesis 4:

 

68

c. Students who were in different sized
high school graduating classes;

d. Students whose fathers and/or mothers
completed grade school, high school,
college, graduate or professional
school;

e. Students who plan to receive one, two,
three or four years of college educa-
tation and those who plan to attend
graduate or professional school;

f. Students whose fathers are executives,
business owners, white-collar workers,
skilled craftsmen, semi-skilled workers,
low or unskilled laborers, farm owners,
public service workers, or professional
personnel (doctor, lawyer, dentist, and
so forth).

There is no relationship between intellec-
tual interest and aptitude scores on the
Scholastic Aptitude Test and Michigan State
University Reading Test, MSU Arithmetic Test,
MSU Mathematics Test and-MSU Quantitative
Test.

 

 

 

 

 

 

There is no relationship between intellec-
tual interest and the following seven
factors of trait self-ratings measured by
the Trait Self-Ratings of College Freshmen:

 

a. Trait—-Physical well-being;

b. Trait--Scholarship;

c. Trait—-Estheticism;
d. Trait-—Pragmatism;
e. Trait—-Technical-scientific;

f. Trait—-Sociability; and
g. Trait—~Sensitivity to others.

There is no relationship between intellec—
tual interest and the following seven
factors of life goals of college freshmen
measured by the Life Goals of College
Freshmen:

 

Hypothesis

 

Hypothesis

 

Hypothesis

 

Hyppthesis

 

Hypothesis

 

69

a. Life goal——Prestige;

b. Life goal-—Personal happiness;

c. Life goa1--Humanistic-cultural;

d. Life goal-~Religious;

e. Life goal—-Scientific;

f. Life goal—~Artistic; and

9. Life goal—~Hedonistic.

There is no relationship between intellec—
tual interest and general self-concept of

academic ability measured by the General
Self-Concept of Academic Ability.

 

The use of a linear model to predict the
MSU grade point average with the predictor
of intellectual interest does not explain
any variance in the criterion variable.

The validity coefficient of intellectual
interest in the prediction of college
success, based on the nonlinear model, is
not statistically significant.

The coefficient of the predictive validity
of intellectual interest with MSU grade
point average as a criterion variable does
not differ whether it is based on the
linear model or on the nonlinear model.

The intellectual interest score does not
improve prediction of the cumulative
college grade point average when it is
added to either of the following:

a. Scholastic Aptitude Test;

 

b. High school grade point average;
0. SAT—Total plus high school GPA;

d. General Self—Concept of Academic
Ability;
e. General Self-Concept of Academic

Ability plus self-reported High
school GPA.

 

 

70

Summary of the Chapter

 

The population under study consisted of all freshmen
who entered Michigan State University Fall term, 1970.
However, students with one or more of the following charac-
teristics were excluded from the population: foreign stu-
dents, transfer students, 1ack of complete data, and Stu—
dents who dropped out before the end of Fall term, 1970.

From this restricted population, 643 students were
randomly selected.

The Academic Interest Scale and MSU Student Survey

 

 

were given to the sample students during the orientation
week, September 21-22, 1970.

The Academic Interest Scale included the Intellectual

 

 

Interest Scale of Stern (Part 1); the Intellectual Interest

 

 

Scale of Anderson and Western (Part 2); and the Intellec-

tualism—Pragmatism Scale of Yuker and Block (Part 3). The

 

MSU Student Survey contained several scales, including the

 

Trait Self-Ratings of College Freshmen (Part A); the Life

 

Goals of College Freshmen (Part B); the General Self-

 

 

Concept of Academic Ability (Part C); and the Student

 

Questionnaire (Part D).

 

Other data, including score on the Scholastic Apti—

 

tude Test, MSU Reading Test, MSU Arithmetic Test, MSU

 

 

Mathematics Test, MSU Quantitative Test, high school grade

 

 

point average, and MSU grade point average were obtained
with the cooperation of the Office of Evaluation Services,

Michigan State University.

71

The criterion for measuring the predictive validity
of intellectual interest was the cumulative college grade
point average subjects obtained at the end of the Fall
term, 1970.

Ten testable statistical null hypotheses were formu-
lated from the purposes of the study and substantive
hypotheses in Chapter I.

As one of the major statistics used in the study, the
principle axis method of factor analysis with the Varimax
rotation was employed to identify the structure and pattern
of intellectual interest. For the purpose of measuring the
validity of intellectual interest in the prediction of
college success, several procedures were used. They
involved testing for linear and nonlinear regression,
plotting the distributions on a scatter-diagram and simple
and multiple regression techniques. The variance—ratio
technique was used to test whether the variable "intellec-
tual interest" improves in accuracy of predicting grade
point average when it is added to the most readily avail-
able predictors.

Chapter IV will deal with the results from analysis

of the data.

CHAPTER IV

ANALYSIS OF THE DATA

This chapter presents the analysis of the data and
the results relating to the main purposes of the study.
The main purposes of the study were, as stated in Chapter
I, to determine the identity and structure of intellectual
interest and, secondly, to test its validity in the pre-
diction of college success in higher education.

Parallel with these purposes, the analysis of data
is presented in three sections: Section 1 presents the
results of factor analysis and the resulting identity and
pattern of the construct of intellectual interest. Section
2 deals with tests of relationship between intellectual
interest as a predictor and the criterion of college
success. Data in regard to validity and reliability are
also presented in Section 2.

Finally, Section 3 deals with the incremental vali-
dity. It indicates whether the test of intellectual
interest improves the predictability of college success
when it is added to other predictors such as results of a
scholastic aptitude test, self-concept of academic ability
and actual and self-reported high school grade point

average.

72

73

Section 1: The Results of Factor Analysis
and Analysis of Variance ‘

 

 

Thirty different variables were employed to test the
statistical hypotheses 1 through 5. The main technique
used was factor analysis. In addition, the analysis of
variance method was also utilized in testing hypotheses
related to the qualitative variables. In this section,
first, null hypotheses which were tested are listed;
second, a description of variables used in the factorial
analytic procedure is presented and, finally, results of the
factorial analysis are reported, including an interpretation
of factored dimensions. Results oftfluaanalysis of variance

are also presented in this section.

A. Hypotheses Tested

 

Hypothesis 1: There is no difference in intellectual
interest, as measured by the Academic
Interest Scale, between males and
females and among:

 

 

a. Students majoring in different
curricula;
b. Students who lived a major portion of

their lives on a farm, in a village,
town, small city, or large city;

c. Students who were in different sized
high school graduating classes;

d. Students whose fathers and/or mothers
completed grade school, high school,
college, graduate or professional
school;

e. Students who plan to receive one, two,
three or four years of college educa—
tion and those who plan to attend
graduate or professional school;

74

f. Students whose fathers are executives,
business owners, white-collar workers,
skilled craftsmen, semi—skilled workers,
low or unskilled laborers, farm owners,
public service workers, or professional
personnel (doctor, lawyer, dentist, and
so forth).

Hypothesis 2: There is no relationship between intellec-
tual interest and aptitude scores on the
Scholastic Aptitude Test and Michigan State
University Reading Test, MSU Arithmetic Test,
MSU Mathematics Test and MSU Quantitative
Test.

 

 

 

 

 

 

 

Hypothesis 3: There is no relationship between intellec-
tual interest and the following seven
factors of trait self-ratings measured by
the Trait Self-Ratings of College Freshmen:

 

 

a. Trait—-Physical well—being;

b. Trait--Scholarship;

c. Trait——Estheticism;

d. Trait—-Pragmatism;

e. Trait-—Technical—scientific;

f. Trait--Sociability; and

g. Trait--Sensitivity to others.
Hypothesis 4: There is no relationship between intellec-

tual interest and the following seven

factors of life goals of college freshmen

measured by the Life Goals of College
Freshmen:

 

 

a. Life goal—~Prestige;

b. Life goal—~Personal happiness;
0. Life goal——Humanistic—cultura1;
d. Life goal—-Re1igious;

e. Life goal~-Scientific;

f. Life goal~~Artistic; and

g. Life goa1--Hedonistic.

Hypothesis 5:

 

75

There is no relationship between intellec-
tual interest and general self-concept of
academic ability measured by the General
Self—Concept of Academic Ability.

 

B. Desctiption of Variables Used in

 

the Factor Analysis Procedure

 

Thirty variables relating to cognitive, affective,

demographic and background characteristics were collected

and used to identify the pattern and structure of the

intellectual

interest trait. The content of each variable

is presented below.

Variable 1:

 

Variable 2:

 

Variable 3:

 

Variable 4:

 

Variable 5:

 

Variable 6:

 

Intellectual Interest
The summated score of 79 items in three sub—
scales of the Academic Interest Scale scored
on the basis of a 2-point scaling system,
as discussed in Chapter III. The score
distribution is shown in Appendix C.

 

Trait Self—Rating--Physical well-being
The content of this variable consisted of
item 1 (Athletic ability), item 8 (Physical
energY) and item 15 (Physical health) of
Trait Self-Ratings of College Freshmen (TSCF),
Part A of the MSU Student Survey.

 

 

Trait Self-Rating—-Scholarship
The content of this variable consisted of
items 2 (Mathematical ability), 9 (Scholar—
ship) and 16 (Intellectual self-confidence)
of the TSCF.

Trait Self-Rating--Estheticism
The content of this variable consisted of
items 3 (Originality), 10 (Artistic ability)
and 17 (Expressiveness) of the TSCF.

Trait Self-Rating--Pragmatism
The content of this variable consisted of
items 4 (Self-control), 11 (Independence)
and 18 (Practical mindedness) of the TSCF.

Trait Self—Rating—-Technical-scientific ability
The content of this variable consisted of
items 5 (Mechanical ability), 12 (Scientific
ability) and 19 (Research ability) of the
TSCF.

Variable

7:

 

Variable

8:

 

Variable 9:

 

Variable

10:

 

Variable

11:

 

Variable

12:

 

Variable

13:

 

Variable

14:

 

76

Trait Self-Rating--Sociability
The content of this variable consisted of
items 6 (Leadership), 13 (Sociability) and
20 (Cheerfulness) of the TSCF.

Trait Self—Rating--Sensitivity to others
The content of this variable consisted of
items 7 (Understanding of others), 14 (Sen-
sitivity to the needs of others) and 21
(Sense of humor) of the TSCF.

Life Goa1-—Prestige
The content of this variable consisted of
items 22 (Becoming a community leader),
29 (Becoming influential in public affairs)
and 36 (Obtaining awards or recognition) of
the Life Goals of College Freshmen (LGCF),
Part B of MSU Student Survey.

 

 

Life Goal--Personal happiness
The content of this variable consisted of
items 23 (Becoming happy and content),
30 (Becoming a mature and well-adjusted
person) and 37 (Becoming a good husband and
wife) of the LGCF.

Life Goal-—Humanistic-cultural
The content of this variable consisted of
items 24 (Developing a meaningful philosophy
of life), 31 (Writing good fiction) and
38 (Keeping up to date with political
affairs) of the LGCF.

Life Goal-~Religious
The content of this variable consisted of
items 25 (Making sacrifice for the sake of
the happiness of others), 32 (Following a
formal religious code) and 39 (Being active
in religious affairs) of the LGCF.

Life Goal--Scientific
The content of this variable consisted of
items 26 (Inventing or developing a useful
product or device), 33 (Making a theoretical
contribution to science) and 40 (Making a
technical contribution to science) of the
LGCF.

Life Goal--Artistic
The content of this variable consisted of
items 27 (Becoming accomplished in one of
the performing arts), 34 (Producing good
artistic work) and 41 (Becoming an accom-
plished musician) of the LGCF.

77

Variable 15: Life Goal-~Hedonistic
The content of this variable consisted of
items 28 (Becoming well—off financially),
35 (Having the time and means to relax and
enjoy life) and 42 (Avoiding hard work) of
the LGCF.

 

Variable 16: General self-concept of academic ability
The content of this variable consisted of
eight items describing different self-
concepts of ability and achievement.

 

Variable 17: Sex -
Male was assigned a score weight of zero
and female a score weight of one.

 

Variable 18: Community lived
Ten categories were in order of urban to
rural area. A weight of zero was given to
"Suburb of a metropolitan area of more than
one million population," 1 to "Suburb of
metropolitan area of 100,000 to 999,999,"
2 to "Suburb of metropolitan area of 25,000
to 99,999," 3 to "In a city (not a suburb)
of more than one million," 4 to "In a city
(not a suburb) of 100,000 to 999,999," 5 to
"In a city (not a suburb) of 25,000 to
99,999," 6 to "In a city of 10,000 to 24,999,"
7 to "In a town of 2,500 to 9.999," 8 to
"In a village of 250 to 2,499," and 9 to
"In a farming or rural community."

 

Variable 19: Father's educational level
Ten categories were made in the order from
lower to higher educational level. A score
weight of zero was given to "Attended grade
school," 1 to "Completed 8th grade," 2 to
"Attended high school," 3 to "Graduated from
high school," 4 to "Technical or business
school beyond high school," 5 to "Attended
college but did not graduate," 6 to "Grad—
uated from college," 7 to "Some education
beyond Bachelor's degree but did not earn
another degree," 8 to "Earned a Master's
degree," and 9 to "Earned a graduate or
professional degree beyond the Master's
level.

 

Variable 20: Mother's educational level
Ten categories which are exactly the same
as for Variable l9.

 

Variable

21:

 

Variable

22:

 

Variable

23:

78

High school size
High school size was classified into five
categories ranging from small to big size
and scored from zero to 4 in that order.

Self-reported high school GPA
The subject was asked to respond in terms
of nine categories ranging from A+ through
C- or lower.

MSU Reading Test.

 

Variable

Variable

24:

25:

MSU Arithmetic Test.

MSU Mathematics Test.

 

Variable

26:

MSU Arithmetic Test plus Mathematics Test.

 

Variable

27:

 

Scholastic Aptitude Test—~Verbal.

 

Variable

28:

Scholastic Aptitude Test-—Mathematics.

 

Variable

29:

Scholastic Aptitude Test--Tota1.

 

Variable

30:

 

Actual high school grade point average.

Mean scores and standard deviation for each of these

thirty variables are shown in Table 4.1.

TABLE 4.l.--Mean Score and Standard Deviation of Each of
the 30 Variables Used in the Factorial Analysis

 

 

Variable Mean Standard Variable Mean Standard
Number Score Deviation Number Score Deviation

l 48.36 10.41 16 31.76 3.57

2 7.44 1.90 17 .51 .50

3 7.61 1.88 18 3.93 3.03

4 6.82 1.74 19 4.46 2.35

5 8.17 1.62 20 3.91 1.75

6 6.61 1.71 21 2.80 1.14

7 7.60 1.76 22 2.90 1.40

8 8.39 1.71 23 33.47 7.15

9 5.43 1.79 24 34.46 4.39

10 10.38 1.56 25 18.19 7.02
11 7.16 1.59 26 52.65 10.59
12 6.30 1.95 27 517.94 114.32
13 4.57 1.95 28 556.32 121.34
14 4.69 1,92 29 1073.28 217.85
15 6.69 1.61 30 3.25 .39

 

79

C. Dimensions Identified through
Factor Analysis

 

 

The intercorrelations of the 30 variables described
above are presented in Appendix D. Principle axis com-
ponents were extracted from the intercorrelation matrix of
these 30 variables and factored into 30 dimensions. The
principle axis method extracts as many factors as variables
entered in the matrix. So, it provided 30 unrotated fac—
tors. These unrotated 30 factors with rounded loadings
for the 30 variables are presented in Appendix E.

The factor eigenvalues are shown in Table 4.2.

TABLE 4.2.--Eigenvalues of 30 Factors Extracted through
Principle Axis Method

 

Factor Factor

 

Number Eigenvalue Number Eigenvalue
1 6.9137 16 .5874
2 3.1134 17 .5529
3 1.9486 18 .5324
4 1.8248 19 .4961
5 1.5425 20 .4888
6 1.3256 21 .4403
7 1.2393 22 .4091
8 1.1725 23 .3994
9 1.0527 24 .3889

10 .9559 25 .3238
11 .7980 26 .2763
12 .7499 27 .2148
13 .7241 28 .1521
14 .6958 29 .0081
15 .6730 30 .0000

 

Ten of the unrotated factors were of acceptable
magnitude with eigenvalues which exceeded .9559. Usually,
an eigenvalue greater than 1.00 is assumed to be acceptable.

According to Table 4.2, factor 10 was short of factor 9 by

80

.0968 and greater than factor 11 by .1579. In other words,
factor 10 had a negligible difference with factor 9 which
exceeded 1.00 in terms of the eigenvalue and a significant
difference from factor 11. So, ten factors were included
in a Varimax procedure.

The ten factors were ranked in terms of their eigen-
values and then rotated; first, the two largest at a time,
then, the three largest, four largest and so on until all
ten were rotated.

The 30 variables with loadings on the ten accepted
factors are shown in Table 4.3 and the 30 variables with
loadings above .50 on the ten factors are shown in Table
4.4.

As noted in Tables 4.3 and 4.4, factor 10 has high
loadings on the intellectual interest variable. The inter-
pretation of factor 10 is directly involved with the test-
ing of hypotheses which were listed in Part A of this
chapter. However, for convenience of presentation, the
labeling and interpretation of the other 9 factors will
precede the testing of the hypotheses.

Each of the individual factors, variable number,
content and loadings exceeding .50 are tabled as a means

of interpretation.

Factor 1: Scholastic eptitude

 

The content of "Factors" was concerned primarily

with scholastic aptitude per se as well as self—rating and

81

.eoppwe° owe ”peace anwooe HH< .

 

 

mac.- mmcl cmc mac .. ecc was: «we 1:3 3
Re eee ewe- eee 8w 3? eee Se SSS-See aw
Se- 8? eee eee- eee Hwe- eee Re Aer-See ew
wed new eee- 3w eee mee- eee E. €1.23 aw
eew- RH... e8.- www- eee e1:- ewe 3e $39 .532 e .efié eee ew
mee- 87 R? eww- eee RT eee eee $39 03222: eee ew
me? eee - 3e 87 eee eee.. Se- eee 389 038132 eee: ew
e3 eww Hee- 1:: eee 3w- Se eee $3. essence eee mw
ewe- ewe e3 wee eee.. wee - ewe- e3 2% me 163831-303 ww
ewe- eee- eee.. eee.. a: Se 8? Se Aowwe Hooeoe eee: Hw
ie- as ewe- ewe Re Se .. eee.. eee :23 .B m . 85.05 ew
Se eee eee.. Se wee eee eee - eww 383 .8 9.35.5 S
ewe eee- wee Rel we see - me e e8 - $2.3 sewfifioee ea
eww emw- emw- eHw ewe mws- eee one - came .3
eew- www eee- wee eee eee.. ewm- ewe Aspwwwpe eo eeooeoe-ewoevmvfi
Re wee - eee.. eee eee.. wwe Se - we? Aoflmwcoeom-ed 3
awe: wa - wwe eR 3e wwe- m we e3 - Sewage-ed 3
eea- eee- eme- eee 38- we eee eee Aoafleowoe-ed Q
eee- eee- eee ewe- eee- eee- wee .. Se- Ameowefioe-ed we
eee- Rew- eee eee eee.. ewe eee.. ewe Aauweeaee-owpmwceaee-eav we
emw see-. mee- eww eee wee eww- ewe Ammocweeae Haeomeoe-eqelefi
e9 - eww- eee «68- see ee- we? wee - Aoewfifiw-Se e
e: 93- eee.. e? 3e .. 8? wt. - ewe- A3218 8. 8838598 e
eee eee.. ewe. eee eee Hew- eee- eee.. Aspwwwpawooe-eem.
eee- eee eme- sew wee eee.. eew- eee flowewpeoaoe-Haoweaooe-eeme
eS- see 38- 3w eee - me e Se - e2 Aauflaefie-S e
wew- flee eee- eee Ree eee- wem- wee Aameowpoepmmieelr
eew- em e ewe .. eee eee .. ewm- eee - wee - Resurgence-S m
eew- eeH- eee 37 e2 eee eee - e8.- QSB-Soz downed-S w
eww-. new eee.. new eee eww- eee.. eew ApmouopeH Heepoeaaoeeu w
. - --, .oz capawwa>
@ m. w W d m. la a .02 nopog

 

.eowpewpa> on one how meeweuoa

popcsom £35 eyepoam ofi popapom .. 38.3.83. soapapom Mesa-fig

I...m.: mama.

82

 

.toppwso wee mpcfi0Q Hesflomp HH< *

 

 

 

-- new -- eee eee em
eee sew eee aw
ewe Hwe- ewe ew
eee eww eee ew
eee wee- eee- ew
ewe sea- Hme- ew
wwe eee- eee ew
eee eew eee mw
eee flee ewe- ww
eee Hefi- wee- aw
eee eee eee ew
ewe eee ewe as
wee eso- woo- efi
eee ewe eee ea
wee mew eefi- ea
eee eee- ewe- es
eee eee mes- ea
eee see eHH- me
eee eww ewe we
eee eee eee- .3
awe eme- mmfi- ea
eee en: eee- m
ewe eee efie- e
are wee emfi- e
eee ewe- mew e
eee eme- wee e
ewe eee eee s
see wee eee- m
eee eee- eee- w
eee eee eew H

.02 menswee>
eww-eqmaaae - New --..-ww--. -

 

 

.ApoBCfipnoov moanwfise> on may pom mmcwtwog
popcsom news whopoum 0H teaspom - memeHac< coepapom xdsfihw>-.m.: mqm<e

 

83

 

.eopewao ore menace stwoee flee .

 

mewe

mmmm

oLr

Neon
Qmmm l

emeel

enopoum

Heme-

swee-
seme-
ewwe
ewme
ewee

owee
mmwet
omen!
onwm .-

NNHN. .-
«ace-

weme.-
afieml.

eeee
eeee-

e e e m. e [-m w

.ucoapapondnoch nopoum one noteboum noegz
oH so on. obona mnacaoq moanaeud> on oopwpom xdsepn>.izeud mAm<e.

mead
meow
oat.
seem
mmeml
mode
emew

Ndmm

mfiwm

em
ew
ew
ew
ew
Aw
ew
mw
ww
aw
ew
2
ea
ew
ea
3
fl
9
we
3
S

HNMd’U‘OB-wo\

.oz canadhd>
woz nope-mm

84

self-concept about scholastic aptitude. It is labeled

"Scholastic aptitude."

TABLE 4.5.--Variable Content of Factor l--Scholastic

 

 

 

 

Aptitude
Vgfiiggie Content Loading
3 Mathematical ability .5815
Scholarship
Intellectual self-confidence
16 Value of self—concept of ability .5242
and achievement
23 MSU Reading Test .6797
24 MSU Arithmetic Test .7963
25 MSU Mathematics Test .7799
26 MSU Qualitative Test .8474
27 SAT—~Verbal .7710
28 SAT—-Mathematics .8975
29 SAT-—Tota1 .9172
Factor 2: Social sensitivity

 

Sociability and sensitive interaction with others are
interpreted as the major components. Physical and practi—
cal self—control had somewhat low negative loadings. The

name "Social sensitivity" is given to this factor.

Factor 3: High school achievement

 

Factor 3 was strongly related to high school grade

point average expressed either in the form of self-report

85

TABLE 4.6.--Variable Content of Factor 2--Social Sensitivity

 

 

Variable .
N ler Content Load1ng

2 Athletic ability -.5960
Physical energy
Physical health

5 Self-control -.6411
Independence
Practical mindedness

7 Leadership .7041
Sociability
Cheerfulness

8 Understanding of others .7122

 

Sensitivity to the needs
of others
Sense of humor

 

TABLE 4.7.--Variable Content of Factor 3--High School

 

 

Achievement
Variable .
N ler Content Load1ng
22 Self-reported high school GPA -.7624
30 Actual high school GPA -.7471

 

or that reported directly by transcripts to the registrar.

The factor is called "High school achievement."

Factor 4: Parents' educational level
Level of education completed by father and mother
have high loadings on factor 4; thus, it is called "Parents'

educational level."

86

TABLE 4.8.—-Variable Content of Factor 4—-Parents' Educa-
tional Level

 

 

Variable -
N ler Content Load1ng
19 Father's educational level .8328
20 Mother's educational level .8517

 

Factor 5: Aesthetic

 

Throughout the six items of the variables, there is
an aesthetic element running through each. This is true
regardless of whether the individual conceives of himself
as "expressive" or as a musical performer. This factor is

labeled "Aesthetic."

TABLE 4.9.--Variable Content of Factor 5-—Aesthetic

 

 

 

 

Variable .
N ler Content Loading

4 Originality .6669
Artistic ability
Expressiveness

14 Becoming accomplished in one of .7780
the performing arts

Producing good artistic work
Becoming an accomplished musician

Factor 6: Community size

 

Community and high school size characterize factor 6.
Community size is described in descending order of large to
small and high school size is described in the opposite

order. The positive direction of community size and the

87

negative direction of high school size actually go in the

same direction. Factor 6 is called "Community size."

TABLE 4.10.--Variable Content of Factor 6--Community Size

 

 

Variable '
N ler Content Load1ng
18 Community size .8025
21 High school size -.7984

 

Factor 7: Conforming-religious

 

Religious confirmity is the major characteristic.
A traditional outlook of happiness and adjustment has some—
what lower loadings on this factor. This factor is labeled

"Conforming-religious."

TABLE 4.1l.-—Variable Content of Factor 7--Conforming-

 

 

 

Religious
Variable .
N ler Content Load1ng
10 Becoming happy and content —.6570
Becoming a mature and well-adjusted
person
Becoming a good husband and wife
12 Making sacrifices for the sake of -.7690
the happiness of others
Following a formal religious code
Being active in religious affairs
Factor 8: Scientific

 

The major loadings of mechanical and scientific

nature 18d t0 naming of factor 8 as "Scientific."

 

88

TABLE 4.12.—-Variab1e Content of Factor 8—-Scientific

 

Variable

Number Content Load1ng

 

6 Mechanical ability -.6842
Scientific ability
Research ability

l3 Inventing or developing a useful -.7899
product or device
Making a theoretical contribution
to science
Making a technical contribution
to science

 

Factor 9: Social hedonism

 

 

Hedonistic context was the largest loading and mascu-
line attitude toward social affairs had the next high

loading on factors. Thus, it is called "Social hedonism."

TABLE 4.13.-—Variable Content of Factor 9--Social Hedonism

 

 

Variable .
N ler Content Loading
9 Becoming a community leader —.6458
Becoming influential in public
affairs
Obtaining awards or recognition
15 Becoming well-off financially -.8237
Having the time and means to relax
and enjoy life
Avoiding hard work
17 Male was assigned a score weight of .5072

zero and female a score weight
of one

 

89

D. The Hypotheses Testing and Definition
of the Construct of Intellectual Interest

 

 

In this section the testing of various hypotheses
generated for this study is discussed in greater detail and
factor 10 is described.

First and foremost, the variable number, name and

loadings on factor 10 are presented in Table 4.14.

TABLE 4.14.-—Variable Content of Factor 10--Intellectual

 

 

Interest
Variable -
N ler Content Load1ng
1 Intellectual interest .5995
11 Life goa1--Humanistic—cultura1 .7675

 

As Tables 4.3, 4.4, and 4.14 indicate, the result of
factor analysis failed to reject Hypothesis 1(b), (c) and
(d) and it was concluded that there was no difference in
intellectual interest between males and females, community
size, high school size and parents' educational level.

Hypothesis 2 was also not rejected since it was found
that there was no relationship between intellectual interest
and the five different aptitude measures used in this study.

Hypothesis 3 was not rejected, i.e., it can be con-
cluded that there was no relationship between intellectual
interest and each of seven factors of trait self-ratings.

All subsets of Hypothesis 4 except (0) were not
rejected and it was concluded that there was no relation-

ship between intellectual interest and these six factors

90

of life goals of college freshmen, namely, prestige, per-
sonal happiness, religious, scientific, artistic and
hedonistic.

Again, Hypothesis 5 was not rejected and it was con-
cluded that there was no relationship between intellectual
interest and general self-concept of ability.

Finally, Hypothesis 4(c) was the only one which was
rejected and it was concluded that not only was there a
relationship between intellectual interest and "Life goal--
Humanistic-cultural," but that both of these constructs
are expressions of the same trait and, furthermore, they
make up a single psychological construct.

The factor analysis indicated that variable 1 (intel-
lectual interest) and variable 11 (Life goal--Humanistic-
cultural) measure the same construct and they make up one
common factor. Therefore, a detailed content of each of
these variables is paramount in understanding the nature
and identity of the factor. In effect, the understanding
of that factor, which is the tenth factor of our factor
analytic procedure, relates to the first purpose of the
study which was aimed at investigating the identify and
structure of the construct of intellectual interest.

A somewhat detailed description of the content of
these two variables is given below.

As explained in Chapter III, this study employed
three subscales to measure the construct of intellectual

interest. Each of the subscales was developed by a

91

different author. Also, these three subscales were assumed
to be based on a similar or very congruent operational
definition of the construct. In addition, it has value to
note that these three subscales evidenced the convergent
validity by showing considerably high intercorrelations.

A summary of the operational definition drawn by each
of the subscales and the content of variable 11 are shown
below.

1. Stern's Intellectual Interest Scale

 

This scale is a subscale of Stern Activities Index

 

and consists of items involving various forms of intellec-
tual activities. These activities are based upon interests
in the arts as well as the sciences, both abstract and
empirical.

2. Anderson and Western's Intellectual Interest Scale

 

Intellectual interest is defined as a dimension of
appreciation and enjoyment of cultural pursuits, and an
interest in philosophical discussion and discourse. Accord—
ing to the authors, there are three aspects to the "Involve-
ment in Intellectual Activity" complex of items of the scale.

The first concerns interest in research and intellec-
tual and academic matters. The second implies an interest
in social and epistemological matters. The third concerns
itself with philosophical and cultural pursuits. In other
words, "Involvement in Intellectual Activity" as assessed

by this scale describes the extent to which individuals

92

enjoy intellectual enquiry and have philosophical and
cultural pursuits (Anderson and Western, 1966; p. 8).

3. Yuker and Block's I-P Scale

 

An intellectualism—pragmatism dichotomy was used not
only in the item content but also in the definitions of
intellectualism and pragmatism. Since, according to the
authors, a pragmatic attitude is essentially anti-intellec-
tual, intellectual attitude is, in turn, anti-pragmatic.

4. Variable 11: Life goal-—Humanistic-cultural

Variable ll consisted of the following items of the

MSU Student Survey: item 24—-"Developing a meaningful

 

philosophy of life;" item 3l—-"Writing good fiction;" and
item 38--"Keeping up to date with political affairs."

A critical examination of these Operational defini-
tions and the content of "Humanistic-cultural life goal"
reveals three essential aspects to the construct of intel—
lectual interest:

The first aspect implies an appreciation and enjoy-
ment of cultural pursuits, the second concerns academic and
philosophical enquiry, and the third aspect concerns anti—
pragmatic interests in the arts as well as in science, both
abstract and empirical.

As indicated above, three hypotheses related to the
qualitative variables were tested with the application of
analysis of variance technique. They were hypotheses 1(a),

1(e) and 1(f).

93

Hypothesis 1(a) stated:

There is no difference in intellectual interest,
as measured by the Academic Interest Scale,
among students majoring in different curricula.

 

Mean and standard deviation of intellectual interest
test score on ten categories of different major fields are
presented in Table 4.15, and the result of analysis of
variance is shown in Table 4.16.

TABLE 4.15.—-Mean Score and Standard Deviation of Intellec-

tual Interest Test Score on Ten Categories of
Major Fields

 

 

. . Standard

Major Fields Mean Deviation
1. Agriculture or Natural Resources 42.15 13.79
2. Arts and Letters 50.14 8.44
3. Business 41.50 8.87
4. Communication Arts 49.70 8.68
5. Education 47.07 9.41
6. Home Economics 40.68 9.23
7. Science or Engineering 50.39 10.40
8. Social Science 49.79 9.87
9. Veterinary or Human Medicine 51.18 9.43
10. No idea what my major field will be 47.97 10.49

 

TABLE 4.16.——Ana1ysis of Variance of the Variable of "Major
Field" with the Dependent Variable of
"Intellectual Interest" Score

 

 

SS d.f. MS F
Major Field 6,287 9 698.56 7.24*
(Between categories)
Error 61,324 633 96.42
(Within categories)
Totals 67,324 642

 

*Significant at d = .01.

94

The test rejected the null hypothesis at the .01
level of significance, and it was concluded that there are
overall differences of intellectual interest scores among
students majoring in different curricula. Major field was
classified into ten categories.

Hypothesis 1(e) is stated as follows:

There is no difference in intellectual interest,
as measured by the Academic Interest Scale, among
students who plan to receive one, two, three or
four years of college education and those who
plan to attend graduate or professional school.

Mean and standard deviation of intellectual interest
test scores on six different categories of educational
expectation are presented in Table 4.17 and the result of
analysis of variance is shown in Table 4.18.

TABLE 4.17.-—Mean Score and Standard Deviation of Intellec-

tual Interest Test Score on Six Categories of
Educational Expectation

 

Educational Expectation Mean StanaFd

eV1ation
l. A year of college 47.00 14.25
2. Two years of college 48.42 6.10
3. Three years of college 48.40 15.50
4. Four years of college 45.56 9.68
5. Master's degree 49.80 9.57
6. Graduate or professional work 51.96 9.97

 

95

TABLE 4.18.—-Analysis of Variance of the Variable of
"Educational Expectation" with the Dependent
Variable of "Intellectual Interest" Score

 

 

 

SS d.f. MS F
Educational Expectation 5,613 5 1,122.71 11.58*
(Between categories)
Error 61,710 637 96.87
(Within categories)
Totals 67,324 642
*Significant at d = .01.

The F-value of the test statistic rejected the null
hypothesis at the d = .01 level of significance, and it
was concluded that there are statistically significant dif-
ferences of intellectual interest scores among the students
with different educational expectation.

Hypothesis 1(f) is stated as follows:

There is no difference in intellectual interest,
as measured by the Academic Interest Scale, among
students whose fathers are executives, business
owners, white-collar workers, skilled craftsmen,
semi—skilled workers, low or unskilled laborers,
farm owners, public service workers, or profes—
sional personnel (doctor, lawyer, dentist, and

so forth).

 

Mean and standard deviation of intellectual interest
test scores on nine categories of fathers' occupations are
presented in Table 4.19 and the result of analysis of

variance is shown in Table 4.20.

 

96

TABLE 4.19.—-Mean Score and Standard Deviation of Intellec-
tual Interest Test Score on Nine Categories of
Father's Occupation

 

 

Father's Occupation Mean Standard

DeV1at1on
1. Semi-skilled worker 45.32 9.90
2. Skilled worker 48.51 9.61
3. Farm owner or operator 44.26 11.71
4. Small business proprietor 47.69 10.82
5. Skilled clerical worker . 45.86 9.90
6. Public service employee 51.70 9.42
7. Executive or managerial 49.35 10.37
8. Professional 49.27 9.77
9. Deceased, retired 51.23 9.23

 

TABLE 4.20.—~Analysis of Variance of the Variable of
Father's Occupation with the Dependent
Variable of Intellectual Interest Score

 

 

SS d.f. MS F
Father's Occupation
(Between Categories) 2,438 8 304.86 2.97*
Error
(Within Categories) 54,885 634 102.86
Totals 57.323 642
*Significant at d = .01.

The test rejected the null hypothesis at the .01
level of significance and it was concluded that there are
statistically significant differences of intellectual
interest according to the different levels of father's
occupation. For the test, the variable of father's occu-

pation was classified into nine categories.

97

Section 2: Reliability, Validity,
and Regression Model

 

 

This section presents the results of hypothesis
testing concerning the validity of intellectual interest in
the prediction of college success. The reliability of the
scale used to measure the trait and the testing of the
regression model are also discussed.

Specifically, Hypothesis6 is related to the validity
of intellectual interest as a predictor of college grade
point average. While Hypothesis 6 is based on the linear
model, Hypothesis 7 is based on the curvilinear model--
although both of them concern predictive validity. Testing
of the appropriateness of either the linear or the curvi—
linear regression model is dealt with in Hypothesis 8.

Yet, before making these hypotheses testings, it
seems to be worthwhile to present information on the reli-
ability of the composite scale used to measure the construct
of intellectual interest.

As indicated by Mehrens and Lehmann (1969, pp. 40-41),
how reliable a test should be in order for it to be useful
cannot be answered in a simple manner. It depends upon the
purposes for which the test is to be used. If it is to be
used to help make decisions about individuals, then it
should be more reliable than if it is to be used to make
decisions about groups of people. Although there is no
universal agreement, it is generally accepted that tests

used to assist in making decisions about individuals should

 

98

have reliability coefficients of at least .80. For group
decisions, a reliability coefficient of about .65 may
suffice.

For the composite, homogeneity of the scale items was
measured by employing the Kuder-Richardson formula 21 (KR
21). Although the formula can be described in several dif—

ferent forms, the following form was used:

 

 

where n represents the number of items in the test,
Mt refers to the mean value of the test, and

SE refers to the variance of the test.

The KR 21 estimate of reliability was .83. Using the
above argument, the reliability coefficient value KR 21 =
.83 indicated that the composite scale was reasonably
reliable.

Hypothesis 6 states that:

The use of a linear model to predict the MSU
grade point average with the predictor of
intellectual interest does not explain any
variance in the criterion variable.

The Pearson product-moment correlation method was
employed to derive the validity coefficient of intellectual
interest in the prediction of college success. The coef-

ficient of validity was .1189 and the coefficient of deter—

mination turned out to be .0141.

 

 

99

To test the statistical significance of the validity
coefficient, the technique of analysis of variance for
overall regression was applied. The result of the analysis
of variance of intellectual interest in predicting college
grade point average is presented in Table 4.21.

TABLE 4.21.-—Ana1ysis of Variance for Overall Regression

of Intellectual Interest with Cumulative
College GPA as a Criterion Variable

 

 

Source of Variance SS d.f. MS F
Regression 467.11 1 467.11 9.19*
Error 3,259.64 641 50.84
Totals 3,726.75 642

*Significant at d = .01.

Consequently, the null Hypothesis 6 was rejected at
the .01 level of significance and the alternative hypo-
thesis was accepted. It was concluded that the predictive
validity of intellectual interest in the prediction of
college success, based on the linear model, was statisti—
cally significant, and, therefore, a linear model did
explain some of the criterion variance.

Since the relation between intellectual interest and
cumulative GPA was found to be statistically significant,
a simple regression equation was set up. Regression coef-
ficients, standard errors of regression coefficients,
standardized beta weights and its standard errors are

presented in Table 4.22.

100

TABLE 4.22.--Regression Coefficient and Its Standard
Errors and Standardized Beta Weights and
Its Standard Errors of Intellectual Interest
with Cumulative College GPA as a Criterion

 

 

Regression Standard Standardized Sgepdard
Coefficients Errors Beta Weights ors
of Betas
Y-intercept 2.3932 0.1336
810pe 0.0082 0.0027 0.12 0 039

 

The least squares regression equation becomes

Y = a + bX
e

where "a" represents the Y—intercept and ”b" the slope of the
line. Both the constants a and b are called regression
coefficient.

Following the information given in Table 4.22, the
least square simple regression equation with intellectual
interest score as a predictor and cumulative college grade
point average as a criterion was found to be

Y = 2.39 + (0.0082) xi

Hypothesis 7 states that:

The validity coefficient of intellectual interest
in the prediction of college success, based on
the nonlinear model, is not statistically sig-
nificant.

The correlation ratio indicated by nyx and fix is a

Y
measure of the relationship which is useful in two circum—

stances:
1. When both variables are continuous but the
regression is not linear, and
2. When one variable is continuous and the other

is discrete.

101

The procedure for computing the value of Eta (ny),
the correlation ration of Y on X, was suggested by Walker
and Lev (1953, pp. 276—278).

The formula suggested by them and used in the study

is as follows:

 

 

 

(T!)2 (T!)2
Z __l__._ __11_
N. N
E2 = 3 = ss Between
YX (T')2 SS Total
, 2
2 N1 (yi) N

where Ni is the total frequency in the i-th row,

n
T! = Z N.. ° y! is the sum of y' scores in the j—th
3 i=1 13 1

column, each multiplied by the appropriate frequency,
and h is the number of rows,
Nj is the total frequency in the j-th column, and

T; = T5 is the sum of the y' score in the j-th column.

With intellectual interest test score as a predictor
and cumulative college GPA as a criterion variable, the
formula was applied after constructing twenty-four cate-
gories for each of the variables. Each category was one-
fourth standard deviation in width. The Eta was found to
be .24.

Furthermore, the test for the null hypothesis nyx = 0

was tested by the F ratio of

 

102

with k representing the number of levels and with

nl = k — l, and n2 = N — k degrees of freedom.

The value of F = 1.52 rejected the null hypothesis
nyx = 0 and it was concluded that the population values
of the correlation ratios was not zero and, in effect, the
validity coefficient of nyx was significantly greater than
zero. '

Hypothesis 8 states that:

The coefficient of the predictive validity of
intellectual interest with MSU grade point
average as a criterion variable does not differ
whether it is based on the linear model or on
the nonlinear model.

Linearity of regression assumes that the relationship
between the criterion and the predictor can be explained by
a linear model. The usual test for linearity of regression
is the test that nyx = pyx' Since both the Pearson product-
moment and the Eta validity coefficient were found to be
statistically significant, it was reasonable to formulate
Hypothesis 8 and test for significance.

The test for linearity of regression is that nyx = pyx
and is made by computing the ratio

 

 

with k representing the number of levels and degrees

of freedom, nl = k - 2 and n2 = N - k.

The value of F = 1.09 failed to reject the null

hypothesis at the .01 level of significance, and it was

103

concluded that the linear model was sufficient. In other
words, the conclusion indicates that, in predicting college
GPA with intellectual interest as a predictor, the curvi—
linear model provides no better prediction than the linear
model.

In addition to the hypothesis testing, a scatter-
diagram is provided in Table 4.23. The table represents
the relationship between the predictor and the criterion
variable. Again, each of these variables is based on
twenty—four categories, each of which was one-fourth of a
standard deviation wide.

Section 3: Incremental Validity of Intellectual
Interest with Cumulative College GPA as a

Criterion and Some Cognitive and
Affective Variable Predictor(s).

 

 

 

 

As Campbell and Fiske (1959) have suggested, any
measuring instrument must show information regarding con-
vergent and discriminant validity. In other words, it is
necessary to demonstrate not only that a measure covaries
with certain other connotatively similar variables, but
also that it's covariance with other connotatively dis—
similar variables is limited.

Yet the most recent developments in measurement
theory suggest that an additional validity evidence should
be presented concerning any test which is intended for pre—
dictive use. Cronbach and Gleser (1957, pp. 30-32) and
Conrad (1950) have both discussed the problem of the base

against which the predictive power of a test is to be

104
Score, Based on 24 Categories, on College

TABLE 4.23.--Scatter Diagram of Intellectual Interest
GPA.

College
GPA

11 3 11 2 1 1

1 1 2 1.212 1
112534. 21224311
221327242331 11
2 13.473133 2 12
5444766‘453211
12323562544313 11
122348348M43 3111
1 334368725251121
1. 252354.5“886521211
1.232333541232211
11.1 Z5536u3m3122
.1724ca124n34=5812199.1. 1;
21.3233 221 1 21..
1 35122 23 a 1

1 22134 2 11
1 1 1 Z 21
1 1 1
1 1
1
1 1

13

.2

to

 

 

2.
11
0.
1

—21

Intellectual
Interest

 

105

evaluated. Cronbach and Gleser declare, "Tests should be
judged on the basis of their contribution over and above

the best strategy available, making use of prior informa-
tion" (1957, p. 31). Such an increase in validity coef-

ficient is called incremental validity.

The incremental validity provides evidence of the
extent to which the test adds to or increases the validity
of predictions made on the basis of data which are already
available.

When a test is added to a battery, the usual way
of expressing its contribution is either as a positive or
a negative improvement to multiple correlation or as no
increment.

The coefficient of correlation between observed scores
on some trait and scores predicted for that trait by a
multiple regression is called a "multiple correlation
coefficient." In a multiple regression equation, scores
on two or more variables will be combined to predict scores
on another variable called the criterion.

Following this line of reasoning, this section
tested and provided the incremental validity or the increase
of predictive validity when intellectual interest variable
is added to one or more of the usually available predictors.

The statistical model employed to test the null hypo—
thesis generated for these purposes was a "variance ratio
test" suggested by Baggaley (1962, p. 21). The statistic

test was:

d.f.

 

where R is the multiple correlation involving m
predictors

R+ is the multiple correlation involving m + l
predictors or m predictor(s) plus intellectual
interest variable,

N is the number of subjects, and

m is the number of predictor(s).

The quotient should be referred to an F table with

l for the numerator mean square and d.f. = N — m — 2

for the denominator mean square.

Hypothesis 9 was generated in the null form, in order

to assess the contribution of intellectual interest to

other predictor(s) in predicting college success in terms

of the cumulative college GPA.

Hypothesis 9 states that:
Intellectual interest score does not improve
prediction of the cumulative college grade
point average when it is added to either of
the following:
a. Scholastic Aptitude Test,
b. High school grade point average,
c. SAT~Total plus high school GPA,
d. General Self—Concept of Academic Ability,

e. General Self-Concept of Academic Ability
plus self—reported high school GPA.

107

The mean score and standard deviation of each of the
predictors used for testing the hypothesis are presented
in Table 4.24, and their intercorrelation coefficients are
shown in Table 4.25.

TABLE 4.24.-—Mean Score and Standard Deviation of the

Criterion Variable of MSU GPA and Five Pre-
dictors Used in the Multiple Regression

 

 

 

 

 

Equation
Standard
Mean . .
DeViation
l. MSU GPA 2.78 .71
2. Intellectual Interest ' 48.36 10.41
3. General Self-Concept of
Academic Ability 31°76 3'57
4. Self-reported high school GPA 7.00 1.40
5. Actual high school GPA 3.25 .39
6. SAT-Total 1,073.28 217.85
TABLE 4.25.--Intercorrelation Coefficient of Criterion
Variable of MSU GPA and the Five Predictors
Used in the Multiple Regression Equation
1 2 3 4 5
l. (MSU GPA)
2. (1.1.) .12
3. (GCAA) .34 .31
4. (SR HS GPA)* .44 .17 .47
5. (HS GPA) .48 .19 .46 .72
6. (SAT-Total) .42 .27 .49 .44 .67

 

*The intercorrelation coefficient value of self—
reported high school grade point average was multiplied
with (-l), because its content in the questionnaire was
written in the format of high value to low value.

108

Hypothesis 9(a) was tested first. The correlation
coefficient involving the predictor of SAT—Total only was
.4217 and the multiple correlation coefficient involving
predictors of SAT-Total and Intellectual Interest test was
.4221. Related results are provided in Table 4.26.

The value of F = 0.31 of the test statistic of a
variance ratio test failed to reject the null hypothesis
9(a), and it was concluded that intellectual interest
score did not significantly increase the predictability of
college success when it was added to SAT—Total as a pre-
dictor.

The correlation coefficient of "High school grade
point average" with college GPA was .4812 and the multiple
correlation coefficient involving HS GPA and the predictor
of intellectual interest was .4820.

The value of the test of a "variance ratio technique"
F = 0.67 failed to reject the null hypothesis 9(b), and
it was concluded that the addition of intellectual interest
test to ”High school grade point average" did not increase
the predictive validity.

The validity coefficient of "SAT-Total and high
school GPA" was .5300 and the addition of the predictor of
intellectual interest to these two predictors made the
validity coefficient .5304.

The F-value<1f0.36 of ”variance ratio method" failed

to reject the null hypothesis 9(c), and it was concluded

109

 

ummumucH HmsuomaamwcH

 

ommv.o mmma.o memo.o mooo.o Hm.m mafia «no mm mm
6:8 amoo .m

. . . . «mo mm mm
mass 0 qmma o memo 0 mm m 8cm 4466 .5

pmmuopsH HmsuomaamucH

somm.o Naoo.o mooo.o oemo.o mH.o- mafia «mo mm
6cm Hmuoeueam .G

. . . . u «no mm
comm o Naoo o mooo 0 mm 0 mafia Hmnoeueam .m

. . . . I ummumucH Hmsuomaamch
ommv o omoo o maoo 0 mo 0 mafia «mm mm .4
mam4.o smoo.o No.0- ssco amo mm .m

. . . . ummumusH HmsuomaamucH
Hmmv o maoo o mooo 0 mm H mafia Hmuoeneam .N
AHm4.o vaoo.o mm.a sfico Hmnoeueam .H

m 9-94m amomm <momm mm «goo .H.H
waspmcoo mcoflpmsqm

 

n a m H o A m u m o o

a o H m m o m m m

 

mSOHumsqm soflmmonmmm mamflpasz 6cm mHQEHm mo musmfloflmmmoo
coflmmwumom cam mucmumcoo paw mpsmHOHmmmoo soauwamuuou oamfluasz Ugo mameflmun.mm.v mqmme

110

that the addition of the predictor of intellectual interest
to "SAT-Total and HS GPA" did not increase the validity.

The F-value of 0.14 failed to reject the null hypo-
thesis 9(d), and it was concluded that the predictor of
intellectual interest did not enhance the prediction of
college success when it was added to the predictor of
"General self-concept of academic ability."

The validity coefficient for predicting college
grade point average involvind the predictors of "General
self-concept of academic ability" and Self-reported high
school grade point average" was .4649, and the addition
of intellectual interest to these two predictors had the
multiple correlation coefficient of .4650.

The F—value of 0.081 failed to reject the null hypo-
thesis 9(e), and it was concluded that the predictor of
intellectual interest did not enhance the prediction of
college success when it was added to "General self-concept
of academic ability" and "Self—reported high school grade
point average."

In short, as shown through the test of hypothesis
9(a) through 9(e), the use of variance ratio test failed
to reject the null hypothesis 9. It was concluded that
the predictor of "intellectual interest" did not improve
the efficiency of the prediction of college success when
it was added to one or more than one of the other four
predictors: Scholastic Aptitude Test, high school grade

point average, general self-concept of academic ability

111

and self-reported high school grade point average. In
other words, the results suggest that the predictor
"intellectual interest" does not have incremental validity

in relation to the above predictors.

Summary of the Chapter

 

The analysis of the data and the results relating
to the main purposes of the study were presented in three
sections.

For the purpose of investigating the identity and
structure of the psychological construct of intellectual
interest, the factor analysis method was employed. In
this procedure, thirty variables concerning the cognitive,
affective and some background characteristics were used.
Principle axis components were extracted from an inter-
correlation matrix of these 30 variables and factored into
30 dimensions. Out of these 30 unrotated factors, ten
factors of acceptable magnitude as indicated by their
eigenvalues were rotated. They were labeled scholastic
aptitude, social sensitivity, high school achievement,
parents' educational level, aesthetic, community size,
conforming—religious, scientific, social hedonism and
intellectual interest.

Three subscales of the composite scale of intellec-
tual interest and the content of the "humanistic—cultural
life goal" formed a common factor which was called "intel—

lectual interest factor."

112

From the critical examination of the operational
definitions drawn in the three subscales of intellectual
interest, which were highly related, and the content of

the "humanistic-cultural life goal,‘ it was suggested that
there were three aSpects of the construct of intellectual
interest. They were an appreciation and enjoyment of
cultural prusuits, academic and philosophical enquiry, and
anti-pragmatic interests in the arts as well as in the
sciences.

The test of analysis of variance indicated that

there were significant differences in intellectual interest,

measured by the Academic Interest Scale, among (1) students

 

majoring in different curricula, (2) students having dif-
ferent levels of educational expectations, and (3) students
whose fathers have different occupations.

The results of the factorial analysis and analysis
of variance were included in Section 1.

Section 2 concerned the validity of intellectual
interest in the prediction of college success with the
criterion of academic grade point average. Also, the
reliability and regression model were tested.

The KR 21 estimate of reliability was .83. It
suggested that the composite scale used to measure the
trait of intellectual interest was very reliable.

The Pearson product—moment correlation coefficient
r = .12 and the correlation ratio n = .24 were obtained.

yx
The former was based on the linear regression model and

113

the latter on the curvilinear model. Both of the predic-
tive validities were found to be statistically significant,
but they were not significantly different. Therefore, the
linear model of regression was adopted and it was con—
cluded that the curvilinear model would provide no better
prediction of college success than the linear model.

Section 3 tested whether or not the addition of
intellectual interest to four other commonly used pre-
dictors increased the predictability of college success.
Very slight gain in multiple correlation could be obtained
by each addition, but none of these increases was statis—
tically significant. Therefore, it was concluded that
intellectual interest had no incremental validity when
added to these predictors.

Chapter V will deal with the summary of the study,
discussion of the results and implications for further

study.

CHAPTER V
SUMMARY, CONCLUSIONS AND

RECOMMENDATIONS

Overview of Purpose and
Instrumentation

 

 

The basic purpose of the study was to determine the
identity and structure of the psychological construct of
intellectual interest and to measure the validity of this
construct in the prediction of the academic success in an
institution of higher learning.

The total population selected for this study con-
sisted of all freshmen entering Michigan State University
in Fall term, 1970. However, students in one or more of
the following categories were excluded from the population:
foreign students, transfer students, students on whom
there was a lack of complete data and students who dropped
out before the end of Fall term, 1970. Five thousand four
hundred and sixty—eight students who registered for credit
courses during the fall registration period were classified
as freshmen.

From this restricted population, 643 students were

randomly selected.

114

115

The Academic Interest Scale and the MSU Student

 

 

Survey were given to the sample of students during the

orientation week, September 21-22, 1970. The Academic

 

Interest Scale included the Intellectual Interest Scale

 

 

of Stern (Part 1), the Intellectual Interest Scale of

 

Anderson and Western (Part 2), and the Intellectualism-

 

Pragmatism Scale of Yuker and Block (Part 3). The MSU

 

Student Survey contained several scales including the

 

Trait Self-Ratings of College Freshmen (Part A), the

 

Life Goals of College Freshmen (Part B), the General Self—

 

 

Concept of Academic Ability (Part C), and the Student

 

Questionnaire (Part D).

 

The other data, such as scores on the Scholastic

 

Aptitude Test, the MSU Reading Test, the MSU Arithmetic

 

 

 

Test and the MSU Mathematics Test, high school grade point

 

average and MSU grade point average, were obtained with
the cooperation of the Office of Evaluation Services,
Michigan State University.

Four subscales of the Stern Activities Index, namely,

 

Reflectiveness, Humanities—Social Sciences, Understanding,

and Science; the Intellectual Interest Scale developed by

 

Anderson and Western; and the Intellectualism-Pragmatism

 

Scale developed by Yuker and Block were employed to
measure the psychological construct of intellectual interest.
All of these subscales were included under the name of the

Academic Interest Scale in order to deter faking responses.

 

The operational definitions of the trait of intellectual

116

interest drawn by each of these three subscales were

similar to each other. The use of three independent

measures also made possible a check on their validity.
The composite score of intellectual interest in the

Academic Interest Scale was derived by applying a 2—point

 

scaling system to all of the 79 items in the three sub-
scales. The equal value was assigned to each of these
items through this method.

The correlation coefficient of the composite score
based on the 2—point scaling system was .89 with the Stern—
total score, .70 with the Anderson and Western-total score
and .63 with the Yuker and Block-total score. These high
intercorrelation coefficients evidenced both the conver—
gent validity of the three subscales of intellectual inter-
est and the justification of the method of the derivation
of the composite score. This manner of deriving a com—
posite score from these three Likert~type scales was
further justified because items in Likert-type scales were
considered to be of approximately equal value.

Data Analysis, Results and
Discussion

 

 

Part A

The first aspect of the purpose of the study was
addressed to reaching a clear understanding of the identity
and structure of the construct of intellectual interest.
This purpose was expressed in the null hypotheses 1 through

5. They are as follows:

Hypothesis 1:

 

Hypothesis 2:

 

Hypothesis 3:

 

117

There is no difference in intellectual inter—
est, as measured by the Academic Interest
Scale, between males and females and among:

 

a. Students majoring in different curricula;

b. Students who lived a major portion of
their lives on a farm, in a village,
town, small city, or large city;

c. Students who were in different sized
high school graduating classes;

d. Students whose fathers and/or mothers
completed grade school, high school,
college, graduate or professional school;

e. Students who plan to receive one, two,
three or four years of college education
and those who plan to attend graduate or
professional school;

f. Students whose fathers are executives,
business owners, white-collar workers,
skilled craftsmen, semi-skilled workers,
low or unskilled laborers, farm owners,
public service workers, or professional
personnel (doctor, lawyer, dentist, and
so forth).

There is no relationship between intellectual
interest and aptitude scores on the Scholas—
tic Aptitude Test and Michigan State Univer-
sitngeadi g Test, MSU Arithmetic Test, MSU
Mathematics Test and MSU Quantitative Test.

 

 

 

 

 

There is no relationship between intellectual
interest and the following seven factors of
Trait Self-Ratings of College Freshmen:

 

a. Trait—-Physical well-being;
b. Trait——Scholarship;

c. Trait—-Estheticism;

d. Trait——Pragmatism;

e. Trait--Technical-scientific;
f. Trait—-Sociability; and

g. Trait—-Sensitivity to others.

118

Hypothesis 4: There is no relationship between intellectual
interest and the following seven factors of
live goals of college freshmen measured by
the Life Goals of College Freshmen:

 

 

a. Life goal—-Perstige;
b. Life goal—-Personal happiness;
c. Life goal--Humanistic-cultural;
d. Life goal-~Religious;
e. Life goal—-Scientific;
f. Life goal-—Artistic; and
g. Life goal-—Hedonistic.
Hypothesis 5: There is no relationship between intellectual
interest and general self-concept of academic

ability measured by the General Self—Concept
of Academic Ability.

 

 

 

These hypotheses were examined through the factor
analytic method. Specifically, the principle axis method
developed by Hotelling (1935) was used. Principle axis
components were extracted from an intercorrelation matrix
of 30 variables related to cognitive, affective, and some
background characteristics. Thirty unrotated factors were
derived.

The following thirty variables were used in the
factorial analysis:

A. The biographical and demographic data

1. Sex
2 Community the student lived most of his/her
life
3. Father's educational level
4. Mother's educational level
5. Size of graduating class of high school
B. The cognitive variables

1. MSU Reading Test

119

MSU Arithmetic Test

MSU Mathematics Test

MSU Quantitative Test

Scholastic Aptitude Test--Verba1

Scholastic Aptitude Test-~Mathematics
Scholastic Aptitude Test-~Total

Self-reported high school grade point average
Actual high school grade point average

KOCDQCNU'lh-WN

C. The affective variables

1. Seven factors of trait self—ratings of
college freshmen

2. Seven factors of life goals of college
freshmen

3. General self—concept of academic ability
4 Intellectual interest

Facilitation of interpretation was further enhanced
by the Varimax method of rotation (1959). Out of 30
unrotated factors, ten factors of acceptable magnitude,
as indicated by their eigenvalues, were rotated. The
purpose of the rotation was to transform the initial
factor solution to achieve simple structure, factor invari-
ance and interpretability. The ten rotated factors were
labeled scholastic aptitude, social sensitivity, high
school achievement, parents' educational level, aesthetic,
community size, conforming-religious, scientific, social
hedonism and intellectual interest.

The result showed that the construct of intellectual
interest had no relationship with the variables of sex,
community size, where student lived most of his/her life,
parents' educational level, size of graduating class of
high school, four kinds of MSU aptitude measures, scholas-
tic aptitude test and high school grade point average.

Also, it had little or no relationship with seven

120

independent factors of trait self-ratings of college
freshmen, general self-concept of academic ability and
with six factors of life goals of college freshmen includ-
ing prestige, prestige, personal happiness, religious,
scientific, artistic and hedonistic.

Three subscales of the composite scale of intellec—
tual interest which had high intercorrelations with each
other and the content of the "humanistic—cultural life
goal" formed a common factor which was called the "intel-
lectual interest factor."

A critical examination of the contents of the items
of the composite scale and the "humanistic—cultural life
goal" suggested three aspects of the construct of intel-
lectual interest. The first aspect implied an appreciation
and enjoyment of cultural pursuit, the second concerned
academic and philosophical enquiry, and the third aspect
concerned anti—pragmatic interests in the arts as well as
in science, both abstract and empirical.

Thus, the person who scores high on the test of this
construct has a high level of interest in ideas, in devel-
opment of philosophical outlook and in what is thought of

H

as "culture. He is also interested in social issues.
Because the construct is not in itself observable, it must
be inferred from the observation of the characteristic

transaction process in which he engages. The evidence in
this study provides the preliminary foundation for further

investigation of the construct and the development of a

reliable and valid measure.

121

For the hypothesis testing related to the demographic
variables which were qualitative, the analysis of variance
method was employed. The analysis of variance test indi-
cated that there were significant differences in intellec-
tual interest, measured by the Academic Interest Scale,
among (1) students majoring in different curricula,

(2) students having different levels of educational expec—
tations, and (3) students whose fathers have different
occupations.

Results of the study of the relationship between the
trait of intellectual interest and major field and level
of educational expectation are especially interesting.

As far as the trait of intellectual interest is concerned,
it is apparent that Michigan State University is no longer,
if it ever was, homogeneous. Students in its different
colleges are apt to differ. To the degree that curricula
are sources of diversity, we would expect that there must
be differential predictors. Whether the construct of
intellectual interest can be established as a predictor

of the field the person will enter is an unanswered
question. However, data in this study do suggest that

there is the possibility such predictions can be made.

Part B

The second aspect of the present study was an attempt
to measure the predictive as well as incremental validity
of intellectual interest in the prediction of college suc—

cess. This effort was reflected in hypotheses 6 through 9.

122

Hypothesis 6: The use of a linear model to predict the
MSU grade point average with the predictor
of intellectual interest does not explain
any variance in the criterion variable.

 

Hypothesis 7: The validity coefficient of intellectual
interest in the prediction of college
success, based on the nonlinear model,
is not statistically significant.

 

Hypothesis 8: The coefficient of the predictive validity
of intellectual interest with MSU grade
point average as a criterion variable does
not differ whether it is based on the
linear model or on the nonlinear model.

 

Hypothesis 9: The intellectual interest score does not
improve prediction of the cumulative col-
lege grade point average when it is added
to either of the following:

 

a. Scholastic Aptitude Test;

 

b. High school grade point average;
c. SAT—Total plus high school GPA;

d. General Self-Concept of Academic Ability;

 

e. General Self-Concept of Academic Ability
plus self—reported high school GPA.

 

Prior to discussing these hypotheses testings, it
seems worthwhile to note again that this study used three
different subscales to measure the psychological construct
of intellectual interest and their high intercorrelations
evidenced their convergent validity. Homogeneity of the
scale items was assessed by employing Kuder-Richardson
formula 21. The KR 21 estimate of reliability was .83,
and it suggested that the composite scale used to measure
the trait of intellectual interest was also very reliable.

The validity of intellectual interest in the pre-

diction of the cumulative college grade point average for

123

the first term at college was determined to be .12, in
terms of the Pearson product-moment method. Also, this
value of the predictive validity was statistically signifi-
cant at d = .01. In addition to that, the validity coef—
ficient nyx based on the curvilinear model was assessed,
and it was found to be .24. This value was also statis-
tically significant at d = .01.

To estimate the regression model between intellectual
interest as a predictor and the first term grade point
average at college as a criterion variable, testing for
linearity of regression was carried out. The test used was
rkwfzpyx' The test concluded that the use of a linear model
in predicting college grade point average, with intellectual
interest as a predictor, would explain a significant amount
of the criterion variance and the use of a curvilinear
model would not reduce the amount of error of prediction
more than the linear mode. A graphical examination pro—
vided the same conclusion.

In practical prediction situations it is seldom that
only one item of information is known about the individual
subject. Therefore, it is valuable to test whether intel—
lectual interest can improve the prediction of college
success when it is added to one or more of the commonly
used academic predictors. This notion was implied in
Hypothesis 9.

When the variable of intellectual interest was added

to a score on the Scholastic Aptitude Test, a slight gain

 

124

of .0004 in multiple correlation could be obtained.
Increases of .0008, .0004, .0003 and .0001 in multiple
correlations were obtained when the test of intellectual
interest was respectively added to high school grade point
average, SAT—Total plus high school grade point average,
the test of general self-concept of academic ability, and
the test of general self-concept of academic ability plus
high school grade point average. None of these additions
of intellectual interest significantly increased the pre-
dictive validity for the first term college grade point
average, and it was concluded that the construct of intel-
lectual interest did not have statistically significant
incremental validity for the predictors investigated.

This finding suggests that, although intellectual interest
is related to a limited extent to college success, it does
not improve the prediction of academic success in college
when it is added to any of four commonly used academic
predictors or typical combinations of them.

The results agree with Goldman (1961) who made the
following observation regarding the relationship between
interests and abilities:

A chronic source of frustration is the failure of
research reports to confirm what seems to be a per-
fectly logical expectation, that it, that interests
and abilities will be closely related, and that
most people will be interested in doing the things
which they are most capable of doing. Counselors
are tempted to see as vexing exceptions those cases

in which interests and abilities point in different
directions. (P. 327.)

125

Yet, the fact seems to be, as concluded by Darley
and Hagenah (1955) after a review of the literature, that
". . . there is a low relation between measured interests
and measured ability and scholastic achievement" (p. 57).

The low correlation and, consequently, small propor—
tion of explained variance might be due to either factors
which interfere with their agreement or due to lack of
true relationship in the data. Several factors seem to
be plausible in the explanation of the present result.

The range of the population upon which the regression
equation is based is restricted to only those who are
admitted to college. The self—selection of students who
decided to apply to Michigan State University rather than
another school also serves to restrict the range. When a
population range is restricted in this manner, the corre-
lation is necessarily reduced.

It is noteworthy that the highest correlations in
other studies were all obtained in Southwestern and Western
colleges in which selective procedures were either so new
or so restricted by statute that they had little effect
on the range of accepted talent (Fishman and Pasanella,
1960). Fishman (1957) also observed a relationship between
scores on college board tests and various characteristics
of students and colleges. The university used in this
study evidently does not represent the spectrum of colleges
in the United States, so general conclusions cannot be

drawn from the results.

126

It is possible to hypothesize that people work harder
on the areas in which they feel weak, sometimes because
they feel challenged, sometimes because it is expected of
them by the elements which influence them such as family
and social values. It is apparent that the findings of
the study leave unanswered the question of what personality
dynamics are operating to produce achievement which is
disparate with our expectation from the intellectual inter—
est score. Even though this study was not planned for
answering that question, it encourages those who seek to
understand the dynamics of achievement through the study
of personality.

In addition, the sample size should be considered.

If the sample size of the groups involved in the study is
large, a statistical test will show a small difference as
significant. In relation to this point, the predictive
validity of intellectual interest seems to be not substan-
tially meaningful even though it was statistically signifi-
cant in this study.

Although grade point average is a socially significant,
objectively quantifiable, and convenient measure of academic
success, it is somewhat inadequate as a single criterion of
college success. It is generally acknowledged that many
kinds of academic learning are not reflected in the GPA.
Measuring success in college with the criterion of grade
point average is not entirely satisfactory. The need for

newer and more relevant criteria is apparent, especially

127

those which offer possibilities for relating the role of
predictors to the specific behaviors contributing to GPA
and the function of the college to the total development
of the individual.

The implications of the study for the college coun-
selor are apparent. One obvious implication is that
incongruency between measured intellectual interest and
academic achievement is to be eXpected in a considerably
large number of cases and is not to be considered as
prima facie evidence of the invalidity of one or both
tests. Another implication is that it may have some value
to be sensitive to the possible presence of one or more
factors emphasized by current learning theories as being
sources of discrepancy between interests and achievement:
family pressures, other needs of the individual which con-
flict with his interests, emotional disturbances, and lack
of experience in utilizing certain abilities of which he

is relatively unaware.

Suggestions for Further Research

 

Based on the information obtained through the review
of the literature, findings of the study and unanswered
questions and possibilities the study has brought into
focus, certain recommendations for further studies can be
made:

1. The present study should be cross-validated to
determine whether the findings when another sample is used

would be congruent with the findings of this study. The

128

population of the present study does not represent the
spectrum of American colleges. Thus, a cross—validational
study based on different populations is encouraged. Also,
the factorial approach employing a considerably larger
number of variables and, as a next step, the known-group
method would be worthwhile to attempt. These studies
might establish foundations for construction of reliable
and valid instruments for measuring the construct as well
as for a clearer understanding of the construct of intel—
lectual interest.

2. Longitudinal studies should be made to include
criteria of college success during the first, second, third,
and fourth years of study at the undergraduate level.
Through extension of the present study, it can be deter-
mined whether the same results found in this study hold
across time. The present study used the first term GPA
only, and it would seem promising to perform follow—up
studies after the students have attended college long
enough for their performance to be adequately evaluated.

3. Studies should be made to discover whether a test
of intellectual interest can be established as a predictor
of the field of study the student will enter. Obviously
the present study did not aim at answering the question.
However, the results of the analysis of variance suggest
a considerable association between the construct and differ—
ent major fields. This finding indicates the possibility

that the test of intellectual interest can serve as a

129

determinant or predictor of the major field of study in
college.

4. Studies should be made to explore the construct
of intellectual interest as a major variable related to
collegiate persistence. It is possible to hypothesize
that dropouts and non-dropouts in college can be charac—
terized by their interest in academic work and related
level of educational expectation.

5. Studies should be made to specify behaviors
contributing to college cumulative grade point average.
Also, specification of the criteria for the total develop-
ment of the individual should be attempted. Then it will
be possible to relate the intellective as well as non-
intellective predictors to these criteria.

6. Studies should be made to determine whether high
school achievement and achievement in college have the same
factors influencing them. Through such studies, explora-
tion of the variables contributing to high school achieve-
ment will also serve in understanding of the dynamics of
college achievement and will be valuable in finding the
predictors of college success.

7. Studies should be made for further pursuit of
possible factors associated with congruence and with dis—
crepancy between an individual's academic interest and
academic achievement. It might be hypothesized, for
example, that correlations between a person's interest and
achievement will be highest in students with the best level

of personality integration.

130

8. Studies should be made to explore whether insti-
tutions and groups of institutions may be differentially
selective or attractive with respect not only to ability,
but also to certain other attributes of personality. It
would be valuable to know what are the interrelationships
among the image or expectations of the university held
by freshmen, the objectives of the institution and the
unique environment of the college. Such information would
be helpful to prospective students as well as college and

university officials.

BIBLIOGRAPHY

131

 

BIBLIOGRAPHY

 

Abelson, R. P. 1952. "Sex Differences and Predictability."
Educational and Psychological Measurement, 12,
638-644.

Aiken, L. R. 1968. "Three Alignment Charts for Use in
Selective Admission." Journal of Educational

 

Research, 62, l4-l7.

 

Anderson, L. B., and Spencer, P. A. 1963. "Personal
Adjustment and Academic Predictability Among
College Freshmen." Journal of Applied Psychology,
47, 97-100.

 

Anderson, D. S., and Western, J. S. 1966. "An Inventory
to Measure Students' Attitudes." University of
Melbourne. Mimeographed.

Asch, S. E. 1952. Social Psychology. New York:
Prentice—Hall.

 

Astin, A. W., and Holland, J. W. 1961. "The Environmental
Assessment Technique: A Way to Measure College
Environments." Journal of Educational Psychology,
52, 308-316.

 

Astin, A. W. 1963. "Further Validation of the Environ-
mental Assessment Technique." Journal of Educa-
tional PsycholQQYr 53, 224-235.

 

 

1965. "An Empirical Characterization of Higher
Educational Institutions.” Journal of Educational
Psychology, 54, 217-226.

 

 

 

 

Atkinson, J. 1958. Motives in Fantasy, Action, and
Society. Edited. New York: Van Nostrad.

Baggaley, Andrew R. 1964. Intermediate Correlation
Methods. New York: John Wiley and Sons, Inc.

 

Baird, L. L. 1969. "Prediction of Accomplishment in
College: A Study of Achievement." Journal of
Counseling Psychology, 16, 246—253.

 

 

132

133

Barclay, J. R. 1965. "Testing for Higher Education."
Student Personnel Series, No. 6. Washington, D.C.:
American College Personnel Association.

 

Barnhart, C. L. 1951. The American College Dictionary.
New York: Random House.

 

Baymur, F. B., and Patterson, C. H. 1960. "A Comparison
of Three Methods of Assisting Underachieving High
School Students." Journal of Counseling Psychology,
7, 83-90.

 

Beatley, B. 1922. "The Relative Standing of Students in
Secondary School in Comprehensive Entrance Exam-
inations and in College." School Review, 30,
141-147.

 

Berdie, B. F. 1961. "Intra-individual Variability and
Predictability." Educational and Psycholpgical
Measurement, 21, 663-676.

 

 

. 1966. "Student Personnel Work: Definition and
Redefinition." Journal of College Student
Personnel, 7, 131-136.

 

 

 

Bishton, R. C. 1966. "Factors Related to the Achievement
of Intellectually Superior Children." Unpublished
Ph.D. dissertation, Ohio State University.

Bloom, B. S., and Peters, F. R. 1961. The Use of Academic
Prediction Scales for Counseling and Selecting
College Entrants. New York: The Free Press of
Glenco, Inc.

 

 

 

Bloom, B. S. 1964. Stability and Change in Human Charac-
teristics. New York: John Wiley and Sons, Inc.

 

1965. "The International Study of Educational
Achievement: Development of Hypotheses." Egg:
ceedings of the 1965 Invitational Conference on
Testing Problems. Educational Testing Service.
106—111.

 

 

Bogue, E. G. 1968. "Application of a Minimum Loss
Decision Strategy in the Selection of Cutoff
Points in College and University Admissions.”
College and University, 43, 131—142.

 

Bolton, C. D., and Kammeyer, K. C. W. 1967. The Univer~
sity Student. New Haven, Connecticut: College
and University Press.

 

 

134

Brewer, J. H. 1963. "An Ecological Study of the Psycho-
logical Environment of Negro College and the
Personality Needs of Its Students." Unpublished
Ph.D. dissertation, University of Texas.

Brookover, W. B., et a1. 1962. Self-Concept of Ability
and School Achievement. U.S. Office of Education
Cooperative Research Project, No. 847.

 

 

. 1965. Self-Concept of Abilipy and School
Achievement. U.S. Office of Education Cooperative
Research Project, No. 1636.

 

 

 

Brown, W. F., and Holtzman, W. H. 1956. Survey of Study
Habits. New York: The Psychological Corporation.

 

Campbell, D. T., and Fiske, D. W. 1959. "Convergent and
Discriminant Validation by the Multitrait-
Multimethod Matrix." Psychological Bulletin, LVI,
81-105.

 

Carter, H. D. 1960. California Study Methods Survgy
(Manual). Los Angeles, California Test Bureau.

 

 

Cattell, R. B., et al. 1962. Prediction and Understanding
of the Effect of Children's Interest Upon School
Performance. U.S. Office of Education Cooperative
Research Project No. 701.

 

 

Centra, J. A. 1967. Student Perceptions of Residence
Hall Environment: Living-Learning_vs. Conventional
Units. Research Memorandum RM—67-l3. Princeton,
New Jersey: Educational Testing Service.

Clark, B. R. 1959. "College Image and Student Rebellion.”
In Selection and Educational Differentiation.
Berkeley, California: Field Service Center and
Center for the Study of Higher Education, Univer-
sity of California.

, and Trow, M. 1966. "The Organizational Context."
In College Peer Groups: Problems and Prospects for
Research. Edited by T. M. Newcomb and E. K. Wilson.
Chicago: Aldine.

 

 

Conrad, H. 1950. "Information Which Should Be Provided
by Test Publishers and Testing Agencies on the
Validity and Use of Their Tests." In The Proceed-
ings of 1949 Invitational Conference on Testing
Problems. Educational Testing Service. 63—68.

 

 

Cosand, J. P. 1953. "Admissions Criteria: A Report of
the California Committee for the Study of Education.‘
College and University, 28, 338-364.

 

135

Cronbach, L. J. 1949. Essentials of Psychological
Testing. New York: Harper anderothers.

 

, and Gleser, G. C. 1957. ngchological Tests
and Personnel Decisions. Urbana: University
of Illinois Press.

 

 

Darley, J. G., and Hagenah, T. 1955. Vocational Interest

 

 

Measurement. Minneapolis: University of Minnesota
Press.

Dohner, W. J. 1969. "The OAIS as Related to Academic
Performance." Journal of College Student

 

Personnel, 10, 44-46.

 

Ellis, A. 1962. Reason and Emotion in Psychotherapy.
New York: Lyle7Stuart.

 

Elton, C. 1969. "Prediction of Educational Outcomes
among Junior College Students." Journal of Collegg
Student Personnel, 10, 44-46.

 

English, H. B., and English, A. C. 1958. A Comprehensive
Dictionary of Psychological and Psychoanalytic
Terms. New York: Longmans.

 

 

Festinger, L. 1957. The Theory of ngnitive Dissonance.

 

 

 

Evanston, Illinois: Row, Peterson.
Fishman, J. A. 1958. "Unsolved Criterion Problems in
the Selection of College Students." Harvard
Educational Review, 28, 320-329.
1960. "An Introduction to the Social—Psychology
of School-to-College Transition.” Journal of

 

Educational Sociology, 33, 249-251.

 

, and Pasanella, A. K. 1960. "College Admission-
Selection Studied." Review of Educational
Research, 30, 298—310.

 

 

Frederiksen, N., and Melville, S. D. 1954. "Differential
Predictability and the Use of Test Scores."

Educational and Psychological Measurement, 14,
647-656.

French, J. W. 1958. "Validation of New Item Types Against
Four-year Academic Criteria." Journal of Educa—
tional Psychology, 21, 250-252.

 

 

Fricke, B. G. 1960. Opinion and Attitude Survey. Evalu-
ation and Examinations Division, University of
Michigan, Ann Arbor.

 

 

136

Garrett, H. F. 1949. "A Review and Interpretation of
Investigation of Factors Related to Scholastic
Success in Colleges and Arts and Sciences and
Teachers Colleges." Journal of Experimental
Education, 18, 91-138.

 

 

Giusti, J. P. 1964. "High School Average as a Predictor
of College Success: A Survey of the Literature."
College and University, 39, 181-194.

 

Goldman, L. 1961. Using Tests in Counseling. New York:
Appleton-Century—Crofts.

 

Golsen, R., et a1. 1960. What College Students Think.
Princeton, New Jersey: H. J. Van Nostrand.

 

Gough, H. G. 1957. California Psychological Inventory.
Palo Alto, California: Consulting Psychologists
Press.

 

Graff, R. W., and Hansen, J. C. 1970. "Relationship of
OAIS Scores to College Achievement and Adjustment."
Journal of College Student Personnel, 11, 129—135.

 

Guilford, J. P. 1962. "Potentiality for Creativity and
Its Measurement." Proceedings of the 1962
Invitational Conference on Testing Problems.
Educational Testing Service, p. 32.

 

Harmon, Harry H. 1960. Modern Factor Analysis. Chicago:
The University of Chicago Press.

 

Harmon, L. R. 1962. "On Decision Making in High School."
National Association of Secondary School Principals
Bulletin, 46, 277.

 

Hatch R. N., Dressel, D. L., and Costar, J. W. 1963.
Guidance Services in the Secondary School.
wm. C. Brown Co.

Havice, C. W. 1966. Campus Values. Boston, Massachusetts:
Northeastern University.

 

Hebenstreit, W. J. 1959. "The Prediction of College Suc—
cess from Biographical Data and Self Ratings."
Mimeographed Report to the Office of Naval Research
for Contract No. Non r-477.

Heckhausen, H. 1967. The Anatomy of Achievement Motiva—
tion. New York: Academic Press.

Heider, F. 1968. The Psychology of Interpersonal Rela-
tions. New York: John Wiley and Sons, Inc.

137

Holland, J. L., and Nichols, R. C. 1964. "Prediction of
Academic and Extra Curricula Achievement in
College." Journal of Educational Psychology, 55,
55-65.

Holland J. L. 1966. "The Prediction of Academic and Non-
academic Accomplishment." Proceedings of the 1966
Invitational Conference on TestingyProblems.
Princeton, New Jersey: Educational Testing Service.

 

 

 

Pp. 45-47.

Hotelling, H. 1935. "Simplified Calculation of Principle
Components." Psychometrika, 1, 27-35.

Hoyt, D. P., and Norman, W. T. 1954. "Adjustment and
Academic Predictability." Journal of Counseling

 

Psychology, 1, 96—99.

 

Jacob, P. E. 1957. Changing Values in College. New York:

 

 

 

Harper.

James, W. 1907. Pragmatism. New York: Green.

Juola, A. E. 1963. "The Development of an Academic Pre-
dictor Scale Based on Students' Attitudes toward
Education." Personnel and Guidance Journal,
381—386.

Kaiser, F. 1959. "Computer Program for Varimax Rotation
in Factor Analysis." Educational and Psychological

 

Measurement, 19, 413-420.

 

Katz, J. 1967. A Portrait of Two Classes: The Under—
graduate Students at Berkeley and Stanford from
Entrance to Exit. U.S. Department of HEW Project
No. 5-0799. Stanford, California: Stanford

 

 

 

 

University.

King, S. H. 1967. "Personality Stability: Early Findings
of the Harvard Student Study." Paper presented
at the American College Personnel Association
Conference.

Klinger, E. 1966. "Fantasy Need Achievement as a Moti—
vational Construct." Psychological Bulletin, 66,
241—308.

Korn, H. A. 1967. Personality Scale Changes from the
Freshmen to the Senior Year. U.S. Department of
HEW Project No. 5-0799. Standord, California:
Stanford University.

 

 

Krathwohl, D. R., et al. 1956. Taxonomy of Educational
Objectives: Handbook II: Affective Domain. New
York: David McKay Co., Inc.

 

 

 

138

Krech, D., et a1. 1962. Individual in Society.
New York: McGraw-Hill Book Company.

 

 

Kurtz, J. J., and Swensen, E. J. 1951. "Factors Related
to Over-achievement and Under-achievement in
School." School Review, 59, 472-480.

Lehmann, I. J., and Dressel, P. L. 1962. Critical

 

Thinking, Attitudes, and Values in Higher Education.
U.S. Department of HEW Project No. 590. East
Lansing, Michigan: Michigan State University.

 

Levin, R. A. 1966. "American College Experience as a
Socialization Process." In College Peer Groups:
Problems and Prospects for Research. Edited by
T. M. Newcomb and E. K. Wilson. Chicago:

 

Aldine.
Linn, R. L. 1966. "Grade Adjustments for Prediction of
Academic Performance: A Review." Journal of

 

Educational Measurement, 3, 313—329.

 

 

Lutz, J. W., and Richards, J. N. 1968. "Predicting
Student Accomplishment in College from the ACT
Assessment." Journal of Educational Measurement,
5, 17—29.

McBee, G., and Duke, R. L. 1960. "Relationship between
Intelligence, Scholastic Motivation, and Academic
Achievement." Psychological Reports, 6, 3-8.

 

McClelland, D., et a1. 1953. The Achievement Motive.

New York: Appleton-Century-Crofts.

 

McConnel, T. R. 1962. "Approaches to the Measurement of
Intellectual Disposition." Proceedings of 1962
Invitational Conference on Testing Problems.
Princeton, New Jersey: Educational Testing
Service, pp. 74-88.

 

McKeachie, W., et a1. 1968. "Student Achievement
Motives, Achievement Cues, and Academic Achieve-
ment." Journal of Consulting and Clinical

 

 

Psychology, 32, 26-29.

Mehrens, W. H., and Lehmann, I. J. 1969. Standardized
Tests in Education. New York: Holt, Rinehart
and Winston, Inc.

 

Merwin, J. C. 1964. "A Nomogram for Assessing Probability
of Success." College and University, 39, 176-180.

139

Morgan, H. H. 1952. "Psychometric Comparison of Achiev-
ing and Non-achieving College Students of High
Ability." Journal of Counseling Psychology, 16,
292-298.

 

Nason, L. J. 1954. "Patterns of Circumstances Related
to Educational Achievement at High School of
Superior Ability." Unpublished Ph.D. dissertation,
University of Southern California.

Newcomb, T. M., et a1. 1967. Persistence and Change:
Bennington College and Its Students after Twenty-
Five Years. New York:. John Wiley and Sons, Inc.

 

 

 

Nicholas, R. C. 1967. "Personality Changes and the
College." American Educational Research Journal,
4, 173-190.

Olson, L. A. 1964. "Students' Reactions to Living,
Learning Residence Halls." Journal of College
Student Personnel, 6, 39-41.

 

Pace, C. R. 1967. Analysis of a National Sample of
Collegg Environment. U.S. Department of HEW
Project No. 5-0764. Los Angeleles, California:
University of California.

 

Paraskevopoulos, J., and Robinson, L. F. 1970. "Compari—
son of Regression Equations Predicting College
Performance from High School Record and Admissions
Test Scores for Males and Females." College and
University, 45, 211-216.

 

 

Pittenger, B. F. 1919. "The Efficiency of College Stu—
dents as Conditioned by Age at Entrance and Size
of High School." The Sixteenth Yearbook of the
National Society for the Study of Education,
Part II. Bloomington, Illinois: Public School

PEBlishing Co.

 

 

Rhine, R. J. 1958. "A Concept Formation Approach to
Attitude Acquisition." Psychological Review,
65, 362-370.

Richards, J. M., Jr. 1966. "A Factor Analytic Study of
the Self-Ratings of College Freshmen." Educa-
tional and Psychological Measurement, 26, 861-870.
1966. "Life Goals of American College Freshmen."
Journal of Counseling Psychology, 13, 12—20.

 

, and Lutz, S. W. 1968. "Predicting Student
Accomplishment in College from the ACT Assessment."
Journal of Educational Measurement, 5, 17-29.

 

140

Rokeach, M. 1960. The Open and Closed Mind. New York:
Basic Books.

 

Rosenberg, M. J. 1956. "Cognitive Structure and Attitu-
dinal Affect." JOurnal of Abnormal and Social
Psychology, 53, 637-672.

 

 

Sanford, M. 1963. "Factors Related to the Effectiveness
of Student Interaction with the College Social
System." In Higher Education and Mental Health.
Edited by B. Berger and E. E. Hall. Gainsville,
Florida: Mental Health Project, University of
Florida. Pp. 8-26.

 

Scheerer, M. 1965. "Cognitive Theory." Chapter 3 of
Handbook of Social Psychology, Vol. I. Cambridge,
Massachusetts: Addison—Wesley.

 

 

Seashore, H. G. 1962. "Women Are More Predictable Than
Men." Journal of Counseling Psychology, 45,
203-210. .

Segal, D. 1934. Prediction of Success in College. U.S.
Department of Interior, Office of Education,
Bulletin No. 18, Washington, D.C.: Government
Printing Office.

 

Shafer, D. 1956. "Analysis of Certain Factors in the
High School Preparation of High School Graduates
Entering Selected Iowa Colleges." Unpublished
Ph.D. dissertation, University of Iowa.

Siegel, M. (Ed.) 1968. The Counseling of College
Students. New York: The Free Press.

Silber, E., et a1. 1961. "Adaptive Behavior in Competent

Adolescent: Coping with the Anticipation of
College." Achievers of General Psychiatry, 5,
354-365.

Smith, L. 1965. "Significant Differences Between High—
Ability Achieving and Nonachieving College Freshmen
as Revealed by Interview Data." Journal of Educa-
tional Research, 59, 10-12. I

 

Spiegel, D. 1971. "Multiple Predictors of Course Grades
for College Men and Women." College Student
Personnel, 12, 44-48.

 

Staton, J. T. 1962. "The Relationship of Selected Factors
to Academic Success for Beginning Freshmen."
Unpublished Ph.D. dissertation, University of
Oklahoma.

 

141

Stebens, L. D. 1957. "A Study of the Relationship Between
Reaking Skills and Academic Achievement in Specific
Subject Matter Areas." Dissertation Abstracts,
17(1), 96-97.

 

Stern G. G. 1963a. Scoring Instructions and College Norms:
Activities Index, College Characteristics Index.
New York: Psychological Research Center.

 

. 1963b. "Characteristics of the Intellectual
Climate in College Environment." Harvard Educa—
tional Review, 33, 5-41.

 

 

 

. 1966. Studies of College Environment. U.S.
Department of HEW Cooperative Research Project No.
378. Syracuse, New York: Syracuse University.

 

 

Strong, E. K. 1943. Vocational Interests of Men and
Women. Stanford, California: Stanford University
Press.

 

Terman, L. M., et a1. 1947. The Gifted Child Grows Up:
Twenty-Five Year Follow-up of Superior Group.
Stanford, California: Stanford University Press.

 

 

Thornburg, L. H. 1924. "College Scholarship and Size of
High School." SChool and Society, 20, 189.

 

Thurstone,L. L. 1947. Multiple Factor Analysis.
Chicago: University of Chicago Press.

 

Travers, R. M. 1949. "Significant Research on the Pre-
diction of Academic Success." In The Measurement
of Student Adjustment and Achievement. Edited
by W. Donahue. Ann Arbor: University of Michigan.

 

 

1959. "The Prediction of Achievement." School
and Society, 70, 293.

 

 

Trent, J. W. 1967. Catholics in College: Religious Com—
mitment and the Intellectual Life. Chicago:
Chicago University Press.

 

Trow, M. 1965. Social Research and Educational Policy:
Research in Higher Education; Guide to Institu-

 

 

tional Decisions. New York: College Examination
Board.

Tyler, R. W. 1962. "The Role of Machines in Educational
Decision-Making." Proceedings of the 1962 Invi—

 

tational Conference on Testing Problem. Princeton,
New Jersey: Educational Testing Service.

142

Wallace, W. L. 1966. Student Culture: Social Structure
and Continuity in a Liberal Arts College. Chicago:
Aldine.

 

 

Ward, A. W. 1959. "The Development and Validation of an
Inventory of Non-cognitive Predictors of Academic
Success." Paper delivered at February, 1959,
Meeting of the American Educational Research
Association.

Webb, 8. 1967. "The Relations of College Grades and
Personal Abilities Considered within Two Frames
of Reference." Multivariate Behavioral Research
Monogrgphs, 67, 53-57.

 

 

Weissman, M. P. 1959. "An Approach to the Assessment of
Intellectual Disposition among Selected High
Ability Students." Unpublished Ph.D. dissertation,
University of California, Berkeley.

Wheeler, S. 1966. "The Structure of Formally Organized
Socialization Setting." In Socialization after
Childhood: Two Essays. Edited by O. G. Brim, Jr.
and S. Wheeler. New York: John Wiley and Sons.

 

 

Weiss, K. P. 1970. "A Multi-factor Admissions Predictive
System." College and University, 45, 203—210.

 

White, H. W. 1951. A Study of the High School Record
as a Criterion for Admission to Purdue University.
Office of Admissions, Purdue University.

Whitla, D. K. 1969. "Research in College Admission."
The Sixty-Eighth Yearbook of the National Society
for the Study of Education. National Society for
the Study of Education. Pp. 81-101.

 

Wrenn, C. G., and Humber, W. J. 1941. "Study Habits
Associated with High and Low Scholarship.”
Journal of Educational Psychology, 32, 611-616.

 

Wrigley, C., and Neuhaus, J. O. 1954. "The Quartimax
Method: An Analytical Approach to Orthogonal
Simple Structure." British Journal of Statistical
Psychology, 7, 81-91.

 

Young, F. C. 1954. "College Freshmen Judge Their Own
Scholastic Promise." Personnel and Guidance
Journal, 32, 399-403.

 

Yuker, H. B., and Block, J. R. 1967. "Intellectual Atti-
tudes and College Performance." Hofstra University
Behavioral Research Information Series, Vol. 1,

No. 1, pp. 1*16.

APPENDICES

 

143

 

APPENDIX A;

 

ACADEMIC INTEREST SCALE

This booklet contains a number of brief statements describing
many different kinds of activities. You will like some of these
things. They will seem more pleasant than unpleasant to you, perhaps
even highly enjoyable. There will be others that you dislike, finding
them more unpleasant than pleasant. The activities listed in this
scale have been obtained from a great many different persons. Peeple
differ in the kinds of things they enjoy, like to do, or find

pleasant to eXperience. You are to decide which of these you like
and which you dislike.

It is hoped that you will give frank and sincere responses.
Our only reason for asking your name and student number is to enable

us to collate the data. Your replies will be held in strict confidence
and will be read only be the research staff.

Do not begin working until you have read, understood and carried
out the directions given.

1. On each of the answer sheets, print in the appropriate places
your name (LAST NAME FIRST) and student number.

2. Unless your STUDENT NUMBER IS CORRECTLY MARKED in pencil in
the rows of Spaces under words STUDENT NUMBER, your answer
sheet CANNOT BE PROCESSED. First, WRITE your student number in
the vertical column of blank boxes under the heavy arrow.
Then, MARK ONE SPACE IN EACH of the SIX ROWS of ten spaces
that corresponds to each number of your student number.

MICHIGAN STATE UNIVERSITY

“"W—mwm' M" swam W m: " '

 

 

 

 

 

 

 

 

 

 

 

...... ....._ ...... .... 5‘ .-1.$1.U 3E A: N I»: BE a; . _.,,
“CTM____INSYRUCTOR j :’:: l 2;; EEE; :3; ::E:: :::.:: :zz'n 2:92: 32::
9'" °' A ... iii: :3: iii :3: :3: .. "'
i BE SURE yous MARKS ARE HEAVY AND LLACK *5 i Q; Q; Q; Q; Q; Q; Q; Q; Q;
; nus: COMPLETELY ANY ANSWER CHARGED :3; .32. 3g; ”:3 A- ....I ”g 31.: ”! 39:)

 

 

' ‘ ' ‘ ....-
.aeeo or... ...-n to... ‘ on... . ...-
. o ....- O...- ..u-. 0.... ....- ...-Q . on... . go... ...-o cu... It... eee-e _-..- . ...-- ..... ..... .
' '9‘..- . on.-- ' on... . ..... . an... 2 ...-on '. .-.-o . an... .......... . ————— ' sense 5... eeeeeeeeeee I
a...
0.... no... ...-0 ...- ..--I . O .
o o 0 o ..... 0.. .-.-u on... ..--c ...-u .0... - ......... ' o-uv. uh... . . . .--.. ...-. ..-” ‘...
5 l:..:: I ::::. 8 .:.:. O ...-. I -..:z Q I ..... I ..... I ..... O ..... I ..... I ..... I ..... ..... Q ..... ..... I... .

\

For each_part of the scale pay particular attention to the

directions. Make sure that you record your responses in the appropriate
spaces on the answer sheet.

page 1

PART 1

The questions on this part are to be answered in the spaces on

your answer sheet numbered 1 through 40.

NH
eee 00

H H u—s Hr—s
C U N HO \OCDVO\\n{-7w
O O O O O O O

...a
U1
0

16.

18.
19.

20,

p
\1
e

 

 

 

Blacken space

1 - if the item describes an activity or event that
you would like, enjoy, or find more pleasant
than unpleasant.

2 - if the item describes an activity or event that
you would dislike, reject, or find more uppleasant
than pleasant.

Igppre columns 31_4 and 5 on the answer sheet.

 

 

Understanding myself better.
Thinking about different kinds of unusual behavior, like
insanity, drug addiction, crime, etc.
Imagining life on other planets.
Trying to figure out why the peeple I know behave the way they do.
Thinking about what the end of the world might be like.
Concentrating intently on a problem.
Finding the meaning of unusual or rarely used words.
Spending my time thinking about and discussing complex problems.
'Working crossword puzzles, figuring out moves in checkers or chess,
playing anagrams or scrabble, etc.
Being a philosopher, scientist, or professor.
Learning how to prepare slides of plant and animal tissues and
making my own studies with a microscope.
Studying wind conditions and changes in atmospheric pressure in
order to better understand and predict the weather.
Reading articles which tell about new scientific developments,
discoveries, or inventions.
Doing experiments in physics, chemistry or biology in order to
test a theory.
Studying the stars and planets and learning to identify them.
Learning about the causes of some of our social and political problems.
Studying the music of particular composers, such as Bach, Beethoven, etc.
Listening to TV or radio programs about political and social problems.
Comparing the problems and conditions of today with those of various
times in the past.
Studying different types of government, such as the A.merican, English,
Russian, German, etc.

PLEASE GO ON TO NEXT PAGE

 

page 2

 

 

21.
22.
23.

24.
25.

26,
27.
28.
29.
30.

31.
32.

33.
34.
35.

38.
39.

Blacken space
1 - if you like,# en 0 or find leasant
2 - if you dislike, re ect or find unpleasant
Ignore columns 3, 4 and 5 on the answer sheet.

 

 

 

Seeking to explain the behavior of people who are emotionally
disturbed.

Reading stories that try to show what people really think and
feel inside themselves.

Imagining what it will be like when rocket ships carry people
through space.

Thinking about the meaning of eternity.

Concentrating so hard on a work of art or music that I don't
know what's going on around me.

Losing myself in hard thought.

Engaging in mental activity.

Solving puzzles that involve numbers or figures.

Following through in the development of a theory, even though
it has no practical applications.

‘working out solutions to complicated problems, even though
the answers may have no apparent, immediate usefulness.

Going to scientific exhibits.

Collecting data and attempting to arrive at general laws
about the physical universe.

Reading scientific theories about the origin of the earth
and other planets.

Reading about how mathematics is used in developing scientific
theories, such as explanations of how the planets move around the sun.
Studying rock formations and learning how they developed.

Talking about music, theatre or other art forms with people
who are interested in them.

Finding out how different language have developed, changed, and
influenced one another.

Learning more about the work of different painters and sculptors.
Studying the development of English or American literature.

Reading editorials or feature articles on major social issues.

PLEASE GO ON TO NEXT PAGE

 

PART 2 page 3

The questions on this part are to be answered in the spaces on
your answer sheet numbered 41 through 42.

 

A Blacken space

- if you rate the statement DEFINITELY TRUE
2 - if you rate the statement MORE TRUE THAN FALSE
3 - if you rate the statement MORE FALSE THAN TRUE
4 — if you rate the statement DEFINITELY UNTRUE
.1 Igpore column 5 on the answer sheet.

 

 

 

41. University subjects which deal with theoretical principles
are generally of greater value to the student than those which
provide information which has a direct practical application.

42. I am more interested in the critical consideration of principles
and theories than in their practical application.

43. I would like to learn more about the history of human thought.

44. I enjoy reading essays on serious or philosophical subjects.

45. I enjoy reading about artistic or literary achievements.

46. I am interested in discussions about such tepics as the ideal
society, freedom, etc.

47. I have spent a lot of time listening to serious music.

48. I like to read poetry.

49. I have frequent discussions with friends about the causes and
possible solutions of various national and international problems.

PLEASE GO ON TO NEXT PAGE

 

 

page 4

PART 2

The questions on this part are to be answered in the spaces on

your answer sheet numbered 93 through 122. Note that you are to
skip the spaces numbered 50 throughpQZ on the answer sheet.

93.

940
95.
96.

98.
99.
100.
101.

102.
103.
104.
105.
106.
107.
108.
109.
110.
111.
112.
113.

114.
115.
116.

117.
118.
119.
120.
121.
122,

 

 

For the statement in each item, blacken space
1 - if you VERY STRONGLY AGREE
2 - if you STRONGLY AGREE
3 - if you AGREE
4 - if you DISAGREE
5 - if you STRONGLY DISAGREE
6 - if you VERY STRONGLY DISAGREE
Ignore columns 7,‘8y,9 and 10 on the answer sheet.

 

 

The primary purpose of higher education should be individual growth
and development.
Americans are too materialistic.
It is important for Americans to understand communism.
Intellectuals should try to be more like normal folks.
Most deep thinkers are too liberal.
Too many gifted peOple have communistic tendencies.
Eggheads get along well with most people.
Ideas are all right, but it's getting the job done that counts.
Colleges and schools should spend more time getting students ready
for jobs and less time filling them with useless information.
Too many new ideas come from radicals and trouble makers.
People with new ideas are usually radicals who are trying to cause trouble.
The greatest contributions to civilization have been made by practical men.
Too few college students are intellectually inclined.
Philosophy is a very valuable study.
Nations are built by hard work, not abstract ideas.
Eggheads should be given more say in politics and government.
PeOple should study many religions before making a choice.
Poetry and art have made many real contributions to civilization.
Bookworms are usually dull peOple.
Being a philosophy major tends to separate one from reality.
If high-brow thinkers would leave well enough alone, we would all
be better off.
Thinkers are more important today than doors.
In today‘s society we need thinkers more than trained personnel.
If I were going to see a play, I would prefer a serious drama
to a musical or light comedy.
A person who goes to a concert is enriching his life.
Music and art courses are usually a waste of time
Most eggheads are snobs.
Colleges should concentrate more on the humanities and less on specialization.
If I had money, I would rather take a trip to Europe than buy a new car.
Artists should go back to painting things as they really are.

- END OF THE SCALE -

 

APPENDIX B .

 

M.S.U. STUDENT SURVEY

Michigan State University would like to know as much as possible
about its students so that it can develop programs and experiences
that will be most beneficial to them. This year, a study is being
made of students entering MSU: their attitudes toward a variety of
subjects, their behavior and their background.

It is hoped that you will give frank and sincere responses.
Our only reason for asking your name and student number is to enable
us to collate the data. Your replies will be held in strict confidence
and will be read only by the research staff.

Do not begin working until you have read, understood and carried
out the directions given.

1. On each of the answer sheets, print in the appropriate places
your name (LAST NAME FIRST) and student number.

2. Unless your STUDENT NUMBER IS CORRECTII’MARKED in pencil in
the rows of spaces under words STUDENT NUMBER, your answer
sheet CANNOT BE PROCESSED. First, WRITE your student number in
the vertical column of blank boxes under the heavy arrow.
Then, MARK ONE SPACE IN EACH of the SI] Rows of ten spaces
that corresponds to each number of'your student number.

IIICIIIGAN STATE WWI!!!"

naeEM—l‘r can ;__nuonn N0. In a I
_

 

 

 

 

 

 

 

 

 

 

 

I I::::: I an: I :2::: e 2:22: I::::: 2 I::::: 'I ::::: I::::: e ::::: I::::x 3|::::: I 333:: I ::::: e um I::::: Q um: I nu: I::::t etzn: ez..::

M.,... W.,... E,— STUDENT NUMBER ;_-
:-l :: :::[:: ngz: ::?:: :13: 31:: :::'z: E: ::!:: :

'CTM_____“IYMYOR Ci: : ’:: .l "9.. ..E... 39:: , ::.’:: I 39:: :3'2: :
ug‘ a 1"? 'u“ :::’:: :::.:: Q , ::?:: "9.. :..-’:: i 329:: :93) :
---( 32:: .:.'.. ..!.. ..E.: ::E:: E i ::T:: ‘:P:: 3.9:: :

I I 3‘“ VOW ”‘RKS ARE “EAVY AND "LACK ’ ‘ :::':: :.!:: :.!:. ::?:: ::..:: ::.':: ‘ ::.’:: ::!:: :zgt: :
' (“ASE CMETELY ANY ANSWER CH‘EMD 39:: :::.:: ::!:z 39:: ‘ 39:: 29:: ::!:: a!" 39:4 :

eee-
eee- so... 0 e on... e e e as. 0 es on e eee eee eeeeeeeeeeeeeeeeeeee . eeeee .... .u-- o .00... 0.... . e
‘ . 0...: . 00-O- . ¢::‘e ‘ cocoa . ...:. ‘ ' ...2: ' :.:.e . :ee:: . .2... . 3.0... I. ..... . ..... . eeeee . eeeeeeeeee l-"J U“:. OI... . .-‘.‘ """

For each part of the survey pay particular attention to the
directions. Make sure that you record your responses in the appropriate
spaces on the answer sheet.

 

 

 

page 1
PART A

The questions on this part are to be answered in the spaces on
your answer sheet numbered 1 through 21.

Read each item carefully; Answer all questions as honestly and
frankly as you can. Only in this way, will the results be meaningful.

 

As you read’each item, mark space
1 - if you rate yourself on the item ggygw AVERAGE
2 - if you rate yourself on the item AVERAGE
3 - if you rate yourself on the item ABOVE AVERAGE
u - if you rate yourself on the item TOP TEN PER CENT

M.
Ignore column 5 on the answer sheet.

Athletic ability
Mathematical ability
Originality
Self-control
Mechanical ability
Leadership
Understanding of others
Physical energy

9. Scholarship

10. Artistic ability

11. Independence

12. Scientific ability

13. Sociability

1“. Sensitivity to the needs of others
15. Physical health

16. Intellectual self-confidence
17. EXpressiveness

18. Practical mindedness
19. Research ability

20. Cheerfulness

21. Sense of humor

 

 

 

mfl O\\n {-‘UNH
0

PLEASE GO ON TO NEXT PAGE

 

page 2
PART B

The questions on this part are to be answered in the spaces on
your answer sheet numbered 22 through #2.

 

As you read each item, blacken space
1 - if you rate the statement OF LITTLE OR NO IMPORTANCE
2 - if you rate the statement SOMEWHAT IMPORTANT
3 - if you rate the statement VERY IMPORTANT
u - if you rate the statement ESSENTIAE FOR YOU

Ignore column 5 on the answer sheet.

 

 

 

 

22. Becoming a community leader.

23. Becoming happy and content.

24. Developing a meaningful philosophy of life.

25. Making sacrifice for the sake of the happiness of others.
26. Inventing or developing a useful product or device.
27. Becoming accomplished in one of the performing arts.
28. Becoming well—off financially.

29. Becoming influential in public affairs.

30. Becoming a mature and we11-adjusted person.

31. Writing good fiction.

32. Followingliaformal religious code.

33. Making a theoretical contribution to science.

34. Producing good artistic work.

35. Having the time and means to relax and enjoy life.
36. Obtaining awards or recognition.

37. Becoming a good husband and wife.

38. Keeping up to date with political affairs.

39. Being active in religious affairs.

#0. Making a technical contribution to science.

41. Becoming an accomplished musician.

32. Avoiding hard work.

PLEASE GO ON TO NEXT PAGE

 

#3.

Q5,

47.

PART C

page 3

The questions on this part are to be answered in the spaces on
your answer sheet numbered 42 through 50.

As you read each item, on your answer sheet,
blacken the number before the statement that best answers

each question.

How do you rate yourself in school
ability compared with your close
friends?

1. I am the poorest

2. I am below average

3. I am average

h. I am above average

5. I am the best

How do you rate yourself in school
ability compared with those in your
class at school?

1. I am among the poorest
2. I am below average

3. I am average

h. I am above average

5. I am among the best

Where do you think you ranked in
your class in high school?

1. among the poorest

2. below average

3. average

a. above average

5. among the best

Do you think you have the ability
to complete college?

1. no

2. probably not

3. not sure either way

a. yes, probably

5. yes, definitely

‘Where do you think you would rank
in your class in college?

1. among the poorest

2. below average

3.

b. above average

average 1

 

5. among the best

48.

In order to become a doctor, lawyer,
or university professor, work beyond
four years of college is necessary.
How likely do you think it is that
you could complete such advanced work?
1. most unlikely

2.mhhm

3. not sure either way

4. somewhat likely

5. very likely

Forget for a moment how others grade
your work. In your own opinion how
good do you think your work is?

1. my'work is much below average

2. my work is below average

3. my work is average

a. my work is good
5. my work is excellent

50.'What kind of grades do you think are

capable of getting?
1. mostly E's
2. mostly 0'8
3. mostly C's
4. mostly 8'8
5. mostly A's

PLEASE GO ON TO NEXT PAGE

 

 

PART D

 

page A

The questions on this part are to be answered in the spaces on
your answer sheet numbered 121 through 132. Note that you are to
skip_the spaces numbered 51 through 120.

As you read each item, on your answer sheet,
blacken the number before the statement that best answers
each question.

121. Sex
1. Male
2. Female

 

122. In which of the broad areas listed below do you expect
to major?

Agriculture or Natural Resources

Arts and Letters

Business

Communication Arts

Education

Home Economics

Science or Engineering

Social Sciences

Veterinary or Human Medicine

I have no idea what my major field will be.

1.
2.
3.
a.
5.
6.
7.
8.
9.
10.

123. Before coming to college, in what kind of a community did you
live most of your life?
1. Suburb of a metrOpolitan area of more than one million
papulation .
Suburb of a metropolitan area of 100,000 to 999,999
Suburb of a metropolitan area of 25,000 to 99,999

2.
3.
4.
5.

7.
8.

9.
10.

In
In
In
In
In
In
In

flhflflbfl.

city (not a suburb) of more than one million
city (not a suburb) of 100,000 to 999,999
city (not a suburb) of 25,000 to 99.999
city of 10,000 to 24.999

town of 2,500 to 9.999

village of 250 to 2.189

farming or rural community

1243‘What kind of high school or secondary school did you attend
(or spend most of your high school years)?
1. Public high school
2. Private, nonreligious, nonmilitary
3. Parochial
a. Others

PLEASE GO ON TO NEXT PAGE

 

 

page 5

125. Highest level of education completed by your father.
(Blacken in ONLY ONE space)
1. Attended grade school (grades 1 to 8 but did not finish)
2. Completed 8th grade
3. Attended high school (grades 9 to 12 but did not graduate)
4. Graduated from high school
5. Technical or business school beyond high school
6. Attended college but did not graduate
7. Graduated from college (Bachelor's Degree)
8. Some education beyond Bachelor's Degree but did not earn
another degree
9. Earned a Master's Degree
10. Earned a graduate or professional degree beyond the
Master's level

126. Highest level of education completed by your mother.
(Blacken in ONLY ONE space)
1. Attended grade school (grades 1 to 8 but did not finish)
2. Completed 8th grade
3. Attended high school (grades 9 to 12 but did not graduate)
h. Graduated from high school
5. Technical or business school beyond high school
6. Attended college but did not graduate
7. Graduated from college (Bachelor's Degree)
8. Some education beyond Bachelor's Degree but did not earn
another degree
9. Earned a Master's Degree
10. Earned a graduate or professional degree beyond the
Master's level

 

127. Father's occupation (select the category that seems to fit best).

1. Semi-skilled worker (e.g., factory worker, service station
attendant)

2. Skilled worker (e.g., plumber, electrician)

3. Farm owner or operator

h. Small business proprietor

5. Office, sales, or skilled clerical worker (e.g., salesman,
bookkeeper, stock clerk)

6. Public service employee

7. Executive or managerial position

8. Professional

9. Deceased, retired

PLEASE GO ON TO NEXT PAGE

 

page 6

128. As you see your situation at the present time, how much education
would you‘like to st?
1. A year of college
2. Two years of college

3. Three years of college

U. Four years of college (Bachelor's Degree)
5. Master's Degree

6. Professional Degree or graduate work beyond Master's level

 

129. Which of the following statements BEST describes you?

(Choose ONE only)

1. I am committed to a particular field of study and am in college
primarily to obtain training for my chosen career.

2. I am primarily concerned with the scholarly pursuit of knowledge,
and the cultivation of the intellect.

3. I believe that the social and extracurricular activities of
college life are just as important as the academic activities. ~-

4. I am committed to a philosophy that emphasized individualistic ?
interests and styles, concern for personal identity, and often,
contempt for many aspects of organised society.

 

130. Of the three purposefully extreme statements, which one comes closest
to describing your parents' policy in regard to your upbringing?
1. All policy in the hands of parents:‘
parents only source of control;
parents domineering and authoritarian
2. Parents suggest without coercing;
parents hope that children will understand reasons for regulations;
parents ready and willing to explain and interpret
3. Great permissiveness: few controls on behavior:
complete freedom for children

131. Size of your high school graduating class
1. Under 25
2e 25 ' 99
3. 100 - 199
u. 200 - 399
5. Over #00

132. Approximately what was your overall grade-point.average in high school?

1. A+

2. A

3. A-

#. 3+

5. B

6. B.

7. C+

8. C

9. C- or lower

- END OF THE SURVEY -

APPENDIX C. Score Distribution of The Academic Interest

 

 

 

Scale

RAW SCORE f of PR gg:score
76 2 2 99 76.9
75 1 3 99 75.9
74 1 4 99 74.0
73 1 5 99 73.0
72 2 7 99 72.0
71 1 8 99 71.1
70 2 10 99 70.1
69 8 18 , 98 69.1
68 6 24 97 68.1
67 5 29 96 67.2
66 7 36 95 66.2
65 14 50 93 65.2
64 12 62 91 64.2
63 14 76 89 63.2
62 12 88 87 62.3
61 9 97 86 61.3 _
6o 13 110 84 60.3 1
59 19 129 81 59.3
58 14 143 79 58.4
57 21 164 76 57.4
55 21 185 73 56.4
55 16 201 70 55.4
54 30 231 66 54.5
53 19 250 63 53.5
52 14 264 60 52.5
51 25 289 57 51.5
50 21 310 53 50.6
49 35 345 49 49.6
48 16 361 45 48.6
47 34 395 41 47.6
46 22 417 37 46.7
85 21 438 34 45.7
44 18 456 30 44.7
43 20 476 28 43,7
42 19 495 24 42.7
41 17 512 22 41.8
40 21 533 19 40.8
39 14 547 16 39.8
38 15 562 14 38,8
37 11 573 12 37.9
36 14 587 10 36.9
35 12 599 8 35.9
34 12 599 8 35.9
33 4 603 7 34.9

 

 

APPENDI§_C. Score Distribution of The Academic Interest
Scale (Continued)

 

 

 

RAW SCORE f of PR z-score
32 7 610 6 34.0
31 10 620 4 33.0
30 7 627 3 32.0
29 4 631 2 31.0
28 1 632 2 30,1
27 3 635 1 29.1
26 3 638 1 28.1
23 2 640 1 25.2
21 1 641 0 23,2
19 1 642 0 21,3
17, 1 643 0 19g3
Mean 48.43

Standard Deviation 10.24
Variance 104.92

 

 
 

 

 

>wmm20HxHu. HademoommeHsdwo: zmdmwx 0% mo <wmwequm smog em «as
meadow >owamHm.

<sHHeGHe no. pl. .0 ? mm 1m. w m

H NHsaoHHoodst Hmwomemdv

m AauwmwuwomH.zeHHIUeusmv ...oow

m AeamorOmemswo .NNV .me

: Aeneassospnausv .mcm .omm .mme

m Aelwmemawdwmav .HO? .NHO .upo .wa

m AelaeowsuoeHlmowosduwwov .NN? .me .tmm .mmm .mwm

V Aeumooppsauaswv .oou .mmo .Ham .mo: .mpm .HHH

m Aaumeampdwdwnw do cavemmv .HHO .Nuu .Hmu .000 .0?» .000 .mpm

e Aboueuamspmov .omo .Hmm .omm .me .oum .omm .mmw .Hms
Ho AbalwommossH msoowaemmv 1.0mo .HHO .owm .omm .HNM .l.owo .mmo .m?m
Hp AbmimssespmauouodecemHv .0NN .omo .HOH .NHm .HHo .oma .H?V .mom
Hm Afimlwewumuoamv 1.00H .000 .000 .omm .OHm .1.oom .Huu .Hmm
H0 AfimnmoeosfiproV .Nu? .om0 .mom .owu .omm .?m0 1.0Hm I.om0
H? AguipudHov . H00 .000 .00? .?0H . 000 .000 .Hom . H?m
H m Abmlmomoapmduov1m0fl.0mm . 0?m 1 . 00H . own 1 . 000 .HO? . 000
pm AmeHwaoomoooa ow.>dHHH¢wv .uom .Hom .mNH .Nuw .mmm .wmm .H0w .H0?
H4 Amoxv .oom .omo .Hse .omm .Hom 1.Hao .owm .omo
pm Aooaasswdw Hw<cdv .1.o#w .000 .00? .I.HH? .OV? Izomm .u.om~ I.HCH
90 Awsareu.m Mm. HodoHV .owm .owv .omH .HOH .oe? .om» .omu .000
we Azossow.m ma. ra4apv .oqq .ouo .ooo .ocm .ow» .oeo .omm .oom
NH 33%» mOWOOH mwsov ... .00N .00? .me .0?N .HHm .0?m . HHN .000
NM AmOthwovolnea mm Dw>v . HVH .owm .?NH .50 . ow» .... .H00 . 0H0 .omw
Nu Azma memdwsm aeuflv .NVH .wa .000 .HHm .000 .Nom 1.ouu 1.00?
N? 350 >HHQ§0¢H0 Homfiv . H00 .OHu .?00 .0H? .HHN .uNN 1 . 02.0 I . 0H0
mm Azma gussaamsuo aamsv .Hmu .ow: .mm» ...omm .omm .6:: \.O6@ -.omm
N0 350 23.9.4.5. Am Sadr. Homdv .Hmm 0.H0 .mflfl I .02. .000 .wmm 1 .00? 1.00?
we Amsaa<v .umw .Hu: .com .Hmo .Ham .mpp 1.Hom .oo:
Nm Am>elzv .HON .0m? .?0m .... .0Nm .00m .00? 1. Hmm 1 . 0m0
N0 Am>aleo¢bHv .Nwm . HHw .?0m 0% . HHO .waH I . pm? I . 0N0
mo “sumo morOOH mw>v .Hmw. .Hm? .ttm omm .ONH mkwmw .vbbwm hbwo

 

——-—-—-——————-—-v1

meadow >3mH%mHm Aoosdwssodv.

>mmeUHM m. szmmoommmHdeos zedwwx 0% mo <pdwsUHmm cmma M: was

 

 

She. 20.10 HP 5. Rm 5 1»? pm pm 3’ H0 H0!
m

w

z.

m

m

w

m

0

Ho. .oou

: .8... L5

Hm .NHm .Nmo .mom

6 . H8 1 .08 . HS . 8»

a: .906 .H::. .Hme .H::. .omo

G 88 .mom 1.8“ :83 .8: .5:

3 .Gm .08 .Hmm .98 b8 .05 n .86

Ho .Hm: .pmu .08» .060 -.Ho: .H:m -.Heo ...Hmm

Hm .omu c.omu -.oou .omp ..o:o -.o:e «.owm 1.0:6 ...owo

S .o:m .88 .05 J8: -.omq 18H ...80 .me {SH .88

8 .86 .08 1.08 1.03 «.08 180 L8 LS 1.84 .88 :8
m» .o:m .98 1.08 .8m .08 :80 .86 .03 .05 1.6mm #5
N» .03 1.8» .08 .8: .80 {Be ...o:m :8 .8: .08 .38
8 .8: 18m .08 {H8 .8: :8: £3: :8 1.03 308 .56
m: .08 380 1.8» .88 ..Sm :56 :5: :8 18:6 .8: be
mm .oom -.om» 1.0m» .omu .mmm -.Heo -.omm .::6 -.mw: 1.046 .Hom
mm .03 -.ofi -85 88 .8» 158 ...HE :8 -.me £08 .HE
we L8 18» .08 £08 .08 :08 {SH :3 :86 no.3 .mom
8 .08 1.8: 1.8... 1.08 L5 :3“ 14$ :5 1.6.8 .8» .3:
mo .cmw .omu .ow: -.ou: .990 -.Hou -.Hmm .:mm -.Hoo ...ou» .How
8 .84 .98» .SH -081 10:8 1.08 URN! 1:8 Foam .081 .o:o

 

 

>‘

 

 

>mmmonxnwb HsdmwooneoHtfiwos zpewwx ow we <swwodwmm dmmm w: firm
m=0dos >3memwm Aoozdwzcmdv.

 

 

 

<5: zo. mo NH 10W 8 m: mm mo no mm mm wo
a

N

w

?

m

a

‘V

m

e

S

3

am

6

a:

3

pm

S

5

pm

8

N» .ome

mu .8: 1:8

8 8:8 .08 -.me

...: .8: .08 L86 :8

8 .08 .08 1.68 .5: 88

8 .08 .68 1:8 :8 88 .68

S 48 .88 :88 .86 :8 :8 .8...

we, .Hmo .omm ..:00 .:m: .80H .uoo .wmo .mmm
8 .Se .8: 1:8 .8: .86 .86 .88 .8: .80
“ma, New» ...on ;.wmw .wmu1 .:~w .:wn .mom 11mwmw 1::0 .:ao

 

 

.11111J1l111111

>wwmzuHN m. mdwmowowo >XHm zwduwx 1 casedwfiod 00 weodomm :00: wounded Howmpsmm
mom 03¢ 00 <emuecwom.

 

 

 

 

0.18.80. 201 H » M : 0 0 0 0 ol HP
<wmweoHe 20.
H 00» 1 »0H 1 :8 1 H 0 0 000 1 8: 10:0 1 H00 H:0 1 H:0
» 1 0»0 1 00: 000 »00 1 H»H 000 0»0 1 H0: 8: 000
0 08 1 »00 H»0 0H0 1 08 1 H:0 H00 00: 1 H8 10:0
: H:0 1 00: 1 0H» 00H »»0 1 0:0 00: H00 . H 00 »0»
0 »H0 1 00: HHH »8 000 .80 »00 H0» H00 1»H0
0 :00 - »0» »00 00: 000 1H00 H0: 1»H0 1H 00 H»H
0 180 1 08 8: H00 1 »00 08 »»» 0»H 0:0 »H0
0 000 1 80 1 000 H»0 1 »»H 1 0H0 H00 H0» »:H 1 0H0
0 1.000 1 :00 00» 1 000 1 0HH 000 1 »00 00» 1 80 .100H
H0 180 1 000 8» 80 1 08 1000 100» 1 000 00: 0»0
HH 00» 1::0 1»H0 1 0H0 HHH »00 1»0» 1H00 000 1H8
H» 1 0:0 1 000 000 1 »0: 1 0H0 1 8H 1000 1:: »»» HOH
H0 »0: 1 H0: 0»0 1 H00 000 1 08 1 H00 1»0» 1&0 B:
H: 1 HOH 1 :00 1 »0: 1 H0: H00 1 8H 1 »00 00: 1 »00 :0»
H0 1 H00 1»H» 000 1 80 1 H»0 H00 1 000 :8 1 000 10H0
H0 000 1 000 00: 1 000 000 000 H8 80 1 HH0 1H 00
H0 1 »»0 100: 1 00H 1 H00 1 »00 10:0 0:0 1 »00 1H0H 000
H0 .1 0:0 H0» 0 00 1 :»» 1 000 :00 00» 1 00: 1 0H: »00
H0 H8 1 80 1 »:0 00H 10 0H :0» 1 0:0 1 »:0 1 »00 1000
»0 H8 1 0»» 1 »:0 :00 1 000 :0H 1 000 10»0 1 » 0: 100:
»H 00: 1 H:H 0:: 000 1 000 1 :H0 1 000 1000 1 0:0 180
»» 08 1 0H0 1 H:0 1 000 1 »0H 1.00: H00 00: 1000 11H 0 0
»0 00: 000 1 »00 0»: 0H0 00» 1000 »:0 H8 80
»: 00: H 00 »»0 000 1 00: 1000 000 10:0 000 HH 0
» 0 08 H8 »00 1 00H 1 H8 1 80 1H»: 1 H8 000 0:H
»0 000 H0» »00 0H0 1 HHH 1 0:0 1 000 1H»0 0:H 000
»0 0:: 000 1 000 00H 0:0 000 1 00: »:0 H:0 0:0
»0 0:0 H8 000 00» 100» 80 1 HHH 000 H00 H»0
»0 00H H0: 1 H 0: 80 1000 00: H8 H00 H 00 H0 0
00 0:0 000 1 H0» 1000 108 ..HOH H00 1 00.0 1 »0: 1»0:

 

1 >HH aeowsew vowswm mum cauddod
I <03p50 space 8.0 uses 5.00 500008" c.

 

 

l i2,

>wmmonx m. wwwsowvwm >xwm Sufiwwx 1 czwo&mdmm mo mmofioam zwwr wocsama bomawzmm
wow 03¢ we <mwwmvwom Aoosdwssmav.

 

, #1....

 

1 HH H» HP H: H0 H0 HN H0 H0 N. - »H
H 1000 000 000 000 1 H00 1H0» H0» H8 1000 1»»0 1 »00
» 1000 11H»: 0»» 1 000 1»0H 10:0 18: 0»: 10»0 00 0 1 00H
0 0H 0 80 1 000 1H00 0»0 H:0 000 1000 1H00 1000 HOH
: »00 1000 1 00: 10:0 00» »H» 1 »»» 000 1»00 0H0 1 »:0
0 1H0: 1H0H 10H0 10:0 1000 1 H00 00» »00 »H0 000 1 00:
0 1»» 0 1 00H 1000 00: 000 »HH H»0 10»0 00» 00H 1 000
0 HHO H00 1 00H 00H 000 »»H HH» 10 0H 000 000 0H0
0 000 HS 1 H00 »0» 00: 1 »00 1 H 00 10:0 8: 000 000
0 H 0 0 1000 1 000 1 000 1 H 00 H00 8: 00» 1 000 10»0 000
H0 1::0 H00 10»0 1 00» HHO HHO 000 100: 1»00 100: 1 00»
HH 1H 0 0 0:0 H»» 10»0 HH: 1 000 1 H00 1»00 0»0 »0H 00»
H» H00 1 :0: 1 000 H»0 10»: 1 0»0 000 1000 H»0 1H:0 150
H0 1»H0 1 0H: 1 H»0 »00 10 00 100: 1000 10:: 0»» 000 00»
H: 00H H00 H00 05 1 0:H 1 H0: »0H 1000 HHO 00» H00
H0 1 »H: HHO 0H0 0:0 1000 1000 1 H 0 0 1 00H H00 08 L00
H0 1 00: 1 00H 000 1 000 0:: 1 0»0 »00 1H0» 1 00» 1. »00 »:0
H0 1H»0 H»0 0H» 1 0»0 1 »»0 H00 1 H3 »00 H00 000 H0:
H0 0»0 H00 1 »0H 1000 1 :00 000 1000 1H 00 1H0: 0:0 00 0
H0 000 1 80 1 0»0 1000 1 000 1 HHH 1.000 1 000 000 100H 0:0
»0 100: 1 000 130 0»H H 00 0H0 00: H8 80 »00 1 00H
»H HH: 10H0 1H 00 1. H00 1:5 10»0 0» 0 1»00 1 H:0 H00 1 0:0
»» 1 000 1H00 000 000 000 1H00 100H 10H0 10:0 000 1H:0
»0 1000 1 000 0:0 000 1000 »00 H»» 1»00 »00 11000 1H0 0
»: H0 0 »0: 0 00 10 00 0 00 00H 1 000 0010 H0» HH» 100H
» 0 H00 H00 1 0H0 1000 0:0 000 10»0 00» 00: 1 000 1 0H0
»0 H0» »HH 8» 1000 00: 0:0 05 0:» 000 00H 1 0:0
»0 1 H00 1H00 000 H00 1. HHH 0»0 10 00 00» 180 SH 000
»0 000 1000 0»0 H» 0 100H 1 00.: 10»: H 0H 1 H:0 000 H00
»0 1 0:0 US 000 H 00 1H0 0 10:0 10:0 H00 1000 000 000
00 10:0 000 1 00N 110%: BL! HH0» 10H0 1 00: 1l»1~1| HH» 1 0001

 

 

 

l , E0:

 

>wwm26Hx m. wawsowowo >me ZhdaHx 1 CSHOdmdwa 0o woodoam zwwr mozzaom Homawsnm
wow 03m wo <demUHom noo:d%:smav.

 

 

 

 

0» 011,1 RF N0 »0 t »w Po
H 1 H00 1 08 100: 0:» 000 0H0 1 000 1 08 1 000
N oHH om: omm oH0 .1om» owo 1oou .1ooH .1ooo
0 1H0» 000 000 000 H0» 0»0 10H: 1 000 1000
: 1 0:0 00H H 00 1000 1000 0010 1 000 00H 1000
0 H00 1000 00H 0»0 1000 0»» 1000 00H 80
0 1000 1 »0H 11000 100H 10»0 1 0»» 0H0 1 00H 08
0 Ham 1m0o Hum omo 1otu 1oom oww 1 ooo .1ooo
0 1H00 »0» 1H00 100H 1000 10»: 100» 100H 000
0 HH» H0» 1H0H 100» 000 100H 100H 80 08
H0 H00 00: 1000 10»0 0»0 00: 000 000 11000
H H 00H 1 0m: 00: 10$ 10H» 10 0H 80 08 1000
H» 1 H00 100: 000 0» 0 000 0H0 .1 000 000 000
H 0 H0 0 »H 0 »»: 0»: 0H0 000 .10» 0 000 1000
H: H»0 10:0 H»0 0:0 10H0 10»0 0H0 80 000
H 0 1 » 00 1 HHH 000 100 0 1 00.0 0» 0 000 000 08
H0 1 08 0»: 000 1 000 10:0 00: 000 000. 00H
H0 1Hou ocm om: .1om0 op» IoHH ..omw ooH ooo
H0 100 0 000 1 000 10»H 000 1 0H 0 0»0 00H 108
Ho Hum .1o::_ 1oNu ooo ouw 1oow vomw ooH ooo
No .1H0m omo omo .1on oom 1oom om» Iooo .1ooo
NH ouo .1oHH 1omm .1omm ouw 1omw ouo ooH .1ooo
mm HHO 1oHu oow Atotp moo 1mm» omw ooo ooo
Nu omo Nwm .1oo0 owe 10H0 1o00 Ipmm VooH ooo
»: 000 8: 100H 1H 00 »0» »8 000 03 1000
»0 1000 00: 0:0 000 1 »:0 1H00 08 100H 1000
»0 8 0 0:0 0»0 1 0» 0 1 00» 10»0 00» 100H 80
»0 0»» 1000 00H 0 00 1 000 How »:0 000 000
mm ..omw “H00 oum 1owm 1000 .Iomw 1mwp 1ocw Iooo
»0 100» H0» 0»0 000 1000 08 10H0 1000 1000
wo omw1 1omp 11om0 war. 1NNo mww 1omo 1oow. vooo