A STUDY OF THE EFFECTS OF GROU? INTERPRETATION OF
APTITUDE TEST RESULTS U?ON THE ESTtMATES OF
ABlUTlES AND THE ESTIMATES OF TEST PERFORMANCE
OF A GROUP OF SEVENTH GRADE PUPELS

Thesis fatheDogmdPhD.
MICHIGAN STATfi UNIVERSITY

C. Lawrence Beymer
1963

 

This is to certify that the

thesis entitled

A STUDY OF THE EFFECTS OF GROUP INTERPRETATION OF
APTITUDE TEST RESULTS UPON THE ESTIMATES OF ABILITIES
AND THE ESTIMATES OF TEST PERFORMANCE OF A GROUP
OF SEVENTH GRADE PUPILS

presented by

Charles Lawrence Beymer

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in Educat ion

     

Major professor

Date

O~169

LIBRARY

Michigan State
University

 

 

OVERDUE FINES ARE 25¢ PER DAY
PER ITEM

Return to book drop to remove
this checkout from your record.

 

 

 

ABSTRACT

A STUDY OF THE EFFECTS OF GROUP INTERPRETATION OF APTITUDE TEST RESULTS
UPON THE ESTIMATES OF ABILITIES AND THE ESTIMATES OF TEST PERFORMANCE
OF A GROUP OF SEVENTH GRADE PUPILS

By C. Lawrence Beymer

This study was designed in order to develop an approach to the
communication and interpretation of test scores and to use the procedure
with a group of seventh grade pupils in order to investigate its
effectiveness. The problem investigated was the effect of group inter-
pretation of aptitude test scores upon the self-estimates of abilities and
the self-estimates of test results of a sample of seventh grade pupils.

The Academic Promise Tests, an aptitude battery measuring abstract
reasoning, numerical, verbal, and language usage abilities, were
administered to seventh grade pupils in one Indiana junior high school.

One month later interpretation sessions were held in regular classroom
groups. After a short discussion of individual differences, abilities,

and norm groups, subjects estimated their abilities in the areas measured
by the tests and their test results. Estimates were made using a chart
composed of numbers and 100 cartoon symbols representing 100 boys and girls
forming a stanine distribution. Group interpretation of test scores
followed immediately, using the test publisher's printed profile. Ten
interpretation groups averaged 28.1 pupils each. One week later subjects
re-estimated their abilities and test results in the same manner as before.

Data for the study were collected from 244 pupils (119 girls and

125 boys) who took the tests and made both preinterpretation and

C. Lawrence Beymer
postinterpretation estimates. Means and standard deviations were computed
for these estimates made by the total group and these sub-groups: boys;
girls; high academic aptitude pupils; low academic aptitude pupils; over-
estimators of abilities; under-estimators of abilities; over-estimators of
test results; and under-estimators of test results. The t-test technique
was used to test the significance of difference between means, the F-test
to test significance of difference between variances, and chi-square to
test the significance of proportions of boys and girls in the various
groups.

Major findings are listed below.

1. Postinterpretation mean ability estimates were significantly
closer to points of maximum estimation accuracy than were preinterpretation
means for these groups: high and low academic aptitude; over-estimators
and under-estimators of abilities and under-estimators of test results.

2. Postinterpretation mean test results estimates were
significantly closer to points of maximum estimation accuracy than were
preinterpretation means for these groups: high and low academic aptitude;
over-estimators and under-estimators of abilities; and over-estimators and
under-estimators of test results.

3. Highly significant decreases in variances of both abilities and
test results estimates were found for the total group and every sub-group
in the study after interpretation.

4. For these groups, variances of postinterpretation estimates of
test results were significantly less than variances of preinterpretation
estimates of abilities; the total study group; boys; girls; high academic

aptitude group; under-estimators of abilities; and under-estimators of

C. Lawrence Beymer
test results.

5. No sex differences in accuracy of estimation were found.

6. Initially, high academic aptitude pupils underestimated, low
academic aptitude pupils overestimated. After interpretation, both groups
moved toward more accurate estimations, the high group close to the points
of maximum estimation accuracy.

7. Postinterpretation estimate means of both under-estimators of
abilities and under-estimators of test results were significantly closer
to the points of maximum estimation accuracy than were postinterpretation
estimate means of over-estimators of abilities and over-estimators of test
results.

8. The high academic aptitude group had significantly less variance
in postinterpretation estimates of abilities and of test results than
did the low academic aptitude group.

9. The under-estimators of abilities and the under-estimators of
test results had significantly less variance in postinterpretation
estimates of test results than did the over-estimators of abilities and

the over-estimators of test results.

All IA“ ii‘.lliill[ii{[1flrlii

Copyright
by
Charles Lawrence Beymer

1961+

A STUDY OF THE EFFECTS OF GROUP INTERPRETATION OF APTITUDE TEST RESULTS
UPON THE ESTIMATES OF ABILITIES AND THE ESTIMATES OF TEST PERFORMANCE

or A GROUP or SEVENTH GRADE PUPILS
By

C. Lawrence Beymer

A THESIS

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

DOCTOR OF PHILOSOPHY
College of Education

1963

ACKNOWLEDGMENTS

No doctoral dissertation has ever been the product of one lone
individual, and this one is no exception. Many people have contributed
much toward its develOpment from genesis to final completion. Thanks
and appreciation are due the following: Dr. John Krumboltz of Stanford
University, for his criticisms and suggestions at the early design stage;
Don Wilkening of the Michigan State University Audio-Visual Department for
his assistance with the stanine distribution chart; Principal Nick Martinez
and his faculty of the waverly Junior High School near Lansing, Michigan,
for permitting numerous try-outs of the interpretation process; Dr. James
Rentschler, Principal of Wbodrow Wilson Junior High School in Terre Haute,
Indiana, and his faculty, who were not only willing and cooperative, but
also eager to have the investigation conducted in their school; Dr. John
Patterson of Purdue University, who contributed several suggestions con-
cerning design and statistical treatment of the data; Dr. Charles Hardaway,
Director of Research and Testing at Indiana State College, Terre Haute,
Indiana, who offered many valuable suggestions during the project; Miss
Janet Stephens, who tabulated the data; Mrs. Janet Kessinger, who typed
and retyped the text several times, and last but not least, my committee,
Dr. James w. Costar, chairman, Dr. Walter F. Johnson, Dr. Gregory A. Miller,
and Dr. Carl F. Frost, who allowed me sufficient freedom and flexibility to
permit the develOpment of a required task into a valuable learning experience.

-CLB

ii

TABLE OF CONTENTS

CHAPTER

I. INTRODUCTION . . . . . . . .

Purposes of the Study. . . . . . .

Guidance Foundations . . . . . . .

The Growth of Standardized Testing

Functions of tests in Guidance and

Counseling.

The Self-theory Frame of Reference . . . .

The Utilization of Test Results. . . . . .

Disuses and Misuses of Standardized Tests.

Statement of the Problem . . . .

SUWIYoooooooooo

II. A REVIEW OF RELATED LITERATURE .

Recent Research

Limitations of previous Studies.

Smryooooooooooooo

III. DESIGN AND METHODOLOGY OF THE STUDY.

ThePrOblemoooooooooooo

Limitations of the Study .
Sample Selection . . . . .
Selection of the Test. . .
Collection of the Data . .
Analysis of the Data . . .
Hypotheses to be Tested. .

sueroooooooooo

r ill

CHAPTER

IV.

FINDINGS OF TIIE STUDY 0 O O O O O O O O O O O O O O O O O O

Study Group Norms Compared to National Norms. . . . . . .

means Of Estimtes. O O O O O O O O O O O O O O O O O O O

Variability of Estimates from Grand Means of Estimates. .

Variability of Estimates from Sex Means . . . . . . . . .

Means of the High and Low Academic Aptitude Groups. . . .

Means of the Over-estimators and Under-estimators of Ability.

Means of the Over-estimators and

Under-estimator of Test

ReSUItSoooooooooooooooooooooooooo

Means of the Sub-groups compared to Grand Means of the Total

StudyGrOLlp...o......o.............

PrOportions of Boys and Girls in the Total Study Group and

intheVariousSub-group8.................

Variability of Estimates of the High and Low Academic Aptitude

PAGE

Groups from the Grand Means of Estimates. . . . . . . . . . . 71

Variability of Estimates of the Over-estimators of Ability and

of the Under-estimators of ability from the Grand Means

OfEStiIDateSooooo00000000000000...

Variability of Estimates of the Over-estimators of Test

Results and of the Under-estimators of Test Results from

the Grand Means of Estimates.

Variability of Estimates of High Academic Aptitude Pupils

Compared to the Variability of Estimates of Low Academic

. 74

O O O O .O O O O O O O O O O O O 77

AptitUdePupi-ISOOOOOOOOOOOOOOOOOOOOOOOBO

iv

CHAPTER PAGE
Variability of Estimates of Over-estimators of Abilities
Compared to the Variability of Estimates of Under-
estimators of Abilities. . . . . . . . . . . . . . . . . . . . 80
Variability of Estimates of Over-estimators of Test Results
Compared to the Variability of Under-estimators of Test
Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Summary of the Findings. . . . . . . . . . . . . . . . . . . . . 86
V. SUMMARY, CONCLUSIONS, RECOMMENDATIONS. . . . . . . . . . . . . . . 91
Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Recommendations for Group Test Results Interpretation. . . . . . 94
Suggestions for Further Research . . . . . . . . . . . . . . . . 95
BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
APPENDIX A. Test Interpretation Script. . . . . . . . . . . . . . . . 103
APPENDIX B. Norm Group Chart in Stanine Form . . . . . . . . . . . . 113
APPENDIX C. Estimation Worksheet . . . . . . . . . . . . . . . . . . 115

APPENDIX DO Fellow-Up ScriptO O O O O O O O O O O O O O O O O O O O O 117

TABLE

I.

II.

III.

IV.

VI.

VII.

VIII.

IX.

LIST OF TABLES

Distribution of Subjects Involved in the Study. . . . .
Median Validity Coefficients Between A.P.T. Scores and
School Grades . . . . . . . . . . . . . . . . . . . .
Reliability Data for the Academic Promise Tests Grade 7
Mean Intercorrelation Coefficients of the A.P.T.. . . .
Academic Promise Tests National Norms Compared to Study
Group Norms . . . . . . . . . . . . . . . . . . . . .
Means of Estimates of Abilities and Positions of Test
Results for Boys, Girls, and Total Group. . . . . . .
Comparison of Means of Preinterpretation and Postinter-
pretation Estimates of Abilities and of Means of
Preinterpretation and Postinterpretation Estimates of
Test Results by Sex and by Total Group. . . . . . . .
Variability of Preinterpretation and Postinterpretation

Estimates of Ability and of Preinterpretation and

Postinterpretation Estimates of Test Results from Grand

PAGE

. 47

Means Of the RespeCtiVe EStimateS o o o o o o o o o o o o o 51

F-Tests of Significance of Variance for Various Combinations

of Total Group Preinterpretation Estimates and

Postinterpretation Estimates of Abilities and Preinter-

pretation and Postinterpretation Estimates of Test Results, 52

vi

TABLE PAGE
X. Variability of Boys' Preinterpretation and Postinterpretation
Estimates of Ability and Preinterpretation and Post-
interpretation Estimates of Test Results from the Grand
Means of Estimates. . . . . . . . . . . . . . . . . . . . . 53
XI. F-Tests of Significance of Variance for Boys' Preinter-
pretation and Postinterpretation Estimates of Abilities
and Preinterpretation and Postinterpretation Estimates
of Test Results from Grand Means of Estimates . . . . . . . 53
XII. Variability of Girls' Preinterpretation and Postinterpretation
Estimates of Ability and Preinterpretation and Postinter-
pretation Estimates of Test Results from the Grand Means
of Estimates. . . . . . . . . . . . . . . . . . . . . . . . 55
XIII. F-Tests of Significance of Variance for Girls' Preinter-
pretation and Postinterpretation Estimates of Abilities
and Preinterpretation and Postinterpretation Estimates of
Test Results from Grand Means of Estimates. . . . . . . . . 55
XIV. F-Tests of Significance of Variance for Various Combinations
of Boys' and Girls' Preinterpretation and Postinterpretation
Estimates of Abilities and Preinterpretation and Postinter-
pretation Estimates of Test Results from Grand Means of
Estimates . . . . . . . . . . . . . . . . . . . . . . . . . 56
XV. Variability of Boys' Preinterpretation and Postinterpretation
Estimates of Ability and Preinterpretation and Postinterpre-
tation Estimates of Test Results from the Means of Boys'

EStthQS-oooooooooooooooooooooooo58

vii

TABLE

XXI.

XXII.

XXIII.

XXIV.

XXVI.

XXVII.

XXVIII.

XXIX.

Comparisons of Means of Preinterpretation and Post-
interpretation Estimates of Abilities and Means of
Preinterpretation and Postinterpretation Estimates
of Test Results of the Over-estimators of Ability
and of the Under-estimators of Ability . . . . . . . . .

Comparisons of Means of Preinterpretation and Post-
interpretation Estimates of Abilities and Preinter-
pretation and Postinterpretation Estimates of Test
results of the Over-Estimators of Test Results and of
The Under-Estimators of Test results . . . . . . . . . .

Comparison of the Means of the High Academic Group With
The Means of the Total Study Group . . . . . . . . . . .

Comparison of the Means of the Low Academic Group With
the Means of the Total Study Group . . . . . . . . . . .

Comparison of the Means of the Over-Estimators of Ability

Group with the Means of the Total Study Group . . . . .

Comparison of the Means of the Under-estimators of Ability
Group with the Means of the Total Study Group. . . . . .

Comparison of the Means of the Over-estimators of Test
Results Group with the Means of the Total Study Group. .

Comparison of the Means of the Under-Estimators of Test
Results Group with the Means of the Total Study Group. .

Chi Square Tests of Significance of the Proportions of

Boys and Girls in the Various Groups of the Study. . . .

ix- xix

PAGE

64

66

67

67

69

69

7O

70

72

TABLE

XXXI.

XXXII.

XXXIII.

XXXIU.

PAGE

Variability of Preinterpretation and Postinterpretation

Estimates of Ability and of Preinterpretation and
Postinterpretation Estimates of Test Results from
Grand Means of the Respective Estimates for the

High Academic Aptitude Group. . . . . . . . . . . . . . . 73

F-Tests of Significance of Variance for Various Combinations

of High Academic Aptitude Groups Preinterpretation and
Postinterpretation Estimates of Abilities and Pre-
interpretation and Postinterpretation Estimates of

Test Results from Grand Means of Estimates. . . . . . . . 73

Variability of Preinterpretation and Postinterpretation

Estimates of Ability and of Preinterpretation and
Postinterpretation Estimates of Test Results from the
Grand Means of the Respective Estimates for the Low

AcademicAptitudeGroup................. 75

F-Tests of Significance of Variance for Various Combinations

of Low Academic Aptitude Groupd Preinterpretation and
Postinterpretation Estimates of Abilities and Preinter-
pretation and Postinterpretation Estimates of Test Results

from Grand Means of Estimates . . . . . . . . . . . . . . 75

Variability of Preinterpretation and Postinterpretation

Estimates of Ability and of Preinterpretation and
Postinterpretation Estimates of Test Results from the Grand

Means of the Respective Estimates for the Over-Estimators

of Ability Group. . . . . . . . . . . . . . . . . . . . . 76

TABLE

XXXVI.

XXXVII.

XXXVIII.

XXXIX.

PAGE

F-Tests of Significance of Variance for Various

Combinations of Over-Estimators of Ability Group's
Preinterpretation and Postinterpretation Estimates
of Abilities and Preinterpretation and Postinterpretation

Estimates of Test Results from Grand Means of Estimates. 76

Variability of Preinterpretation and Postinterpretation

Estimates of Ability and of Preinterpretation and
Postinterpretation Estimates of Test Results from the
Grand Means of the Respective Estimates for the Under-

estimators of Ability Group. . . . . . . . . . . . . . . 78

F-Tests of Significance of Variance of Various Combinations

of Under-Estimators of Ability Group's Preinterpretation
and Postinterpretation Estimates of Abilities and
Preinterpretation and Postinterpretation Estimates of

Test Results from Grand Means of Estimates . . . . . . . 78

Variability of Preinterpretation and Postinterpretation

Estimates of Ability and of Preinterpretation and
Postinterpretation Estimates of Test Results from the
Grand Means of the Respective Estimates for the Over-

estimators of Test Results Group . . . . . . . . . . . . 79

F-Tests of Significance of Variance for Various Combinations

of Over-estimators of Test Results Group's Preinterpretation
and Postinterpretation Estimates of Abilities and
Preinterpretation and Postinterpretation Estimates of

Test Results from Grand Means of Estimates . . . . . . . 79

xxi

TABLE

XXXXI.

XXXXII.

XXXXIII.

XXXXIV.

PAGE

Variability of Preinterpretation and Postinterpretation

Estimates of Ability and of Preinterpretation and
Postinterpretation Estimates of Test Results from the
Grand Means of the Respective Estimates for the Under-

eStimAtorS Of TESt RESUICB GrOUp o o o o o o o o o o o o 81

F-Tests of Significance of Variance for Various

Combinations of Under-estimators of Test Results
Group's Preinterpretation and Postinterpretation
Estimates of Abilities and Preinterpretation and
Postinterpretation Estimates of Test Results from Grand

Means Of Estimates 0 o o o o o o o o o o o o o o o o o o 81

F-Tests of Significance of Variance for Preinterpretation

and Postinterpretation Estimates of Abilities for
Preinterpretation and Postinterpretation Estimates of
Test Results Comparing High Academic Aptitude Pupils

With LOW Academic Aptitude Pupils. o o o o o o o o o o o 82

F-Tests of Significance of Variance for Preinterpretation

and Postinterpretation Estimates of Abilities and For
Preinterpretation and Postinterpretation Estimates of
Test Results Comparing the Over-estimators of Ability

With the under-€8t1mat0t8 Of Ability o o o o o o o o o o 84

P-Tests of Significance of Variance for Preinterpretation

and Postinterpretation Estimates of Abilities and for
Preinterpretation and Postinterpretation Estimates of
Test Results Comparing Over-estimators of Test Results

with Under-estimators of Test Results. . . . . . . . . . 85

xxii

CHAPTER I

INTRODUCTION

I. PURPOSES OF THE STUDY

The purposes of this study were: (1) to develop an approach to
the communication and interpretation of test scores, and (2) to use
the procedure with a group of seventh grade pupils in order to inves-

tigate its effectiveness.

II. GUIDANCE FOUNDATIONS

Practically non-existent until the turn of this century, guidance
and counseling in the schools has become established as one of the basic
components of our educational system. Because a wide variety and com-
bination of activities have evolved to meet local needs, attempting to
generalize about guidance and counseling practices is difficult. However,
even though specific practices and organizational patterns vary from
school to school and from state to state, the goals are usually the same.
Simply stated, they are: to assist the individual student to understand
himself, his environment, and to make wise decisions in light of these
understandings. A "wise choice" is defined as one the individual makes
for himself after considering the relevant factors and alternatives, a
choice for which he is willing to take the consequences. This is the

philosophical framework for the present study.

III. THE GROWTH OF STANDARDIZED TESTING

Despite the above~mentioned diversity in pattern and practice
psychological testing is undoubtedly one of the most widespread and
universal components of guidance programs in our schools. For docu-
mentation of this contention, there are figures from many sources which
reveal the great number of standardized tests and answer sheets purchased
each year. It has been estimated that in 1944 more than 26 million tests
were administered by educational institutions, business firms, and personnel
consultants to more than 11 million individuals.1 Another source2
estimates 1954 sales at 61 mdllion and the number for the 1958-1959
school year at 122 million. In the academic year beginning in September,
1961, and running through June, 1962, the estimated number of standardized
tests administered in schools had risen to 125 million, a figure just
shy of three times the total number of all students enrolled in public
and private education from kindergarten through graduate school.3 The
.1221 _A_n_n_ua__l_ m g_f_ Mardized Test Publishers prepared by Stanley B.

Hunt and Associates for the American Textbook Publishers Institute4 in

 

1"Testing is Big Business," American Psychologist, II, January,
1947, p. 26.

2"Testing: Can Everyone Be Pigeonholed?" Newsweek, 54, 3,
July 20’ 1959, pp. 91-93.

39Are Americans Over-Testing?" Overview, August, 1961, pp. 31-33.
41961 Annual Survey g; Standardized Test Publisherg, prepared by

Stanley B. Hunt and Associates for the American Textbook Publishers
Institute, 432 Park Avenue South, New York City, New York, April, 1962.

3
April, 1962, contains the following statistics on net sales of standardized

tests and answer sheets.

1954 ‘ 81,526,000
1955 83,800,000
1956 91,070,000
1957 97,810,000
1958 109,710,000
1959 133,620,000
1960 - Total 140,750,000
Grades 1 - 8 81,800,000
Grades 9 - 12 40,850,000
Grades 13 & over 12,400,000
Industrial 5,700,000
1961 - Total 141,100,000
Grades 1 - 8 81,850,000
Grades 9 - 12 41,950,000
Grades 13 and over 11,550,000
Industrial 5,750,000

Although figures from.these sources differ slightly, it seems obvious
that testing with standardized instruments has become a popular practice
in our schools. Gibson5 administered questionnaires to 904 high school
students in 12 schools of three states. Of this number, only Eggs;
individuals reported that they had not taken such tests.

Hardaway, Rozak, and Ederle6 provide the most recent survey data

on the extent and cost of testing in the secondary schools of one state,

 

5Robert L. Gibson, "Pupil Opinions of High School Guidance
Programs," Personnel 53g Guidance Journal, XL, 5, January, 1962,
Pp s 453.457.

6Charles W. Bardaway, Grace 3. Kozak, and Helen Ederle, Budgetagy
gg§.Pinancial Provisions 32; Guidance Services igHIgdiana Publig.8econda£z
Schools, Office of Research and Testing, Indiana State College, Terre
Baute, Indiana, 1963.

4
Indiana. From 661 questionnaires mailed out, 323 usable returns were
received, representing 452.0f all secondary schools, public and private,
in the state. A further breakdown revealed that 101 of the state's small
schools (enrollment, less than 100), 462.0f the medium-sized schools
(enrollment, 100 to 249), and 782 of the large schools (enrollment 250

or more) responded. Data pertinent to this study are as follows:

small medium large total

 

Number reporting use of tests
(2.0f schools responding) 732 831 911 862

Annual per pupil expenditure
for standardized tests 93¢ 66c 98c 95¢

IV. FUNCTIONS OF TESTS IN GUIDANCE AND COUNSELING

Information that standardized tests can supply have many possible
uses in the school. Thorndike and Hagen7 suggest three kinds of functions:
classroom, administrative and guidance. Examples of classroom functions
include identifying pupils who need special diagnostic study and evaluating
discrepancies between potentiality and achievement after determining
reasonable achievement levels for each pupil.Typical administrative
functions are forming and assigning to classroom groups, and evaluating
.curricular offerings, emphases, and experiments. Of special interest

to this study are the guidance functions of standardized tests which

emphasize helping pupils set educational and vocational goals, make

 

7Robert L. Thorndike and Elizabeth Hagen, Measurement and
Evaluation.ig Psychology and Education, Second Edition (New York:
Wiley 8 Sons, 1961), p. 446.

5
immediate choices, and build realistic self-pictures. Thus test results
have the potentiality for facilitating progress toward the important goal
'mentioned earlier: helping individuals make wise decisions through the
use of information about themselves and their environments. Such a
function rests squarely upon an acknowledgment of the importance of an

individual's concept of self as a determinant of his behavior.
V. THE SELF-THEORY FRAME OF REFERENCE

Self-theorists consider an individual's self-concept to be of
crucial importance. Combs and Snygg8 state that what a person thinks
and how he behaves are largely determined by his conceptions of him-
self, including his abilities. Bordin9 says that the individual responds
to choices in terms of his conception of himself, and Superlo uses the
implementation of the self-concept as the framework for his theory of
occupational choice and vocational development.

It is generally agreed that this concept of self is not inborn,

11

but must be achieved. Rogers says that the self and personality

 

QArthur Combs and D. Snygg, Individual Behavior, Revised Edition,
(New York: Harper & Brothers, 1959).

9B. S. Bordin, “A Theory of vocational Interests as Dynamic

Phenomena," ‘ggggggign.1.ggg Psychological Measurement, 3, 1943,
pp. 49-65.

10Donald B. Super, "vocational Adjustment: Implementing a Self-
Concept,“ Occupations, XXX, Rovember, 1951.

1l'Association for Supervision and Curriculum Development, National
Education Association, Perceiving, Behaving, Becomigg (Washington, D.C. :
I. B. A. ’ 1962) O

6
emerge from experience. Jersild12 writes that the self-concept develops
as the person meets up with the experiences of life. This point of view

13 and Combslh.

is supported by White

Several authorities have taken the position that the establishment
of the self-concept is one of the major tasks of the adolescent period.
Erikson15 considers the establishment of identity as the main deve10p-
mental task of adolescence. Priedenberg16 says that the central growth
process in this age period is to define the self through clarification
of experience and to establish self-esteem.

If the self-concept is of crucial importance in determining
behavior, if it is learned from experience, and if much of this con-
struction occurs during adolescence, then several implications for
educational practice can be identified. Jersild has written that:

..... the most important task for child psychology and for

education is to find out how the educational program.from

nursery school through college might help the growing person
to understand and accept himself.17

 

12Arthur T. Jersild, .15 Search 2; Self, (New*York: Bureau of

Publications, Teachers College, Columbia University, 1952).

13Robert H. white, The Abnormal Personality (New York: The
Ronald Press, 1956).

14Arthur H. Combs, "A Perceptual View of the Adequate Personality,"
Perceivigg, Behavigg, Becoming (Washington, D. C.: N. E. A., 1962).

158. H. Brikson, Childhood 59g Society (New‘York: Norton, 1955).

16Edgar z. Friedenberg, The vanishing Adolescent (New’York:
Dell Publishing Company, 1962).

l‘7Arthur T. Jersild, in Search 9; Self (New York: Bureau of
Publications, Teachers College, Columbia University, 1952), p. 3.

7
Peters and Farwell18 suggest that perhaps the most important discovery
that boys and girls need is to make a discovery of themselves. Super
supports this view specifically: "Here, then, is a major goal for
education: the deve10pment of clear, well-formulated, and realistic
self-concepts."19 The guidance services, a part of the total educational
program, share this responsibility. Priedenberg states: "The purpose
of guidance, after all, is to help students see themselves clearly and
realistically, and to accept what they see at least as good enough to
go on from."20

It appears that several specific kinds of things can be done to
help students develop realistic self-concepts.

First of all, we can direct our emphasis toward the assets of
each individual instead of toward his liabilities. Combs21 says that
people learn that they are able not from failure, but from success.
The writers of the _]_._9_6_2_ _A_. g. _c_. Q. Yearbookzz point out that often
it is the child who feels unwanted and unable who cannot afford to

be accurate in self-assessment.

 

18Herman J. Peters and Gail P. Parwell, Guidance: A Bevelomnental
Approach (Chicago: Rand HcNally & Company, 1959).

19Donald E. Super, The Psychology g£_Careers (New Ybrk: Harper
6 Brothers, 1959), p. 111.

20Edgar z. Priedenberg, Th5,Vanishigg.Adolescent (New'YOrk:
Dell Publishing Company, 1962), p. 144.

21Arthur H. Combs, “A Perceptual View of the Adequate Personality,"
Perceiving, Behavin , Becoming (washington, D. C.: N.E.A., 1962), pp. 50-64.

22.Association for Supervision and Curriculum Development, National

Education Association. Perceivigg, Behavigg, Becoming (washington, D.C.:
N. E. A., 1962), p. 120.

8

Second, if we really want people to develop unique, personal
self-concepts, we must not only accept but encourage variability.

Third, the individual must be accepted by others, including
his teachers and counselors, if he is to learn to accept himself.
Rogers and his co-workers have demonstrated that being accepted leads
to self-acceptance and that, in turn, leads to acceptance of others.

Fourth, we should make available to each student all of the
valid and reliable information about himself that he needs in order to
make decisions about his own future. This includes the results from

standardized tests, the particular focus of this study.
VI. THE UTILIZATION or TEST RESULTS

While some may believe that test results are for the private
use of the expert and for no other eyes, most authorities disagree with
this proposition. The authors of the 1962 A. _S_. g. 2. Yearbook take the
following position.
Any information, including achievement and intelligence test
ratings, which contributes to the accuracy of his view of self
should be available to the student. Withholding records of
performance, test results, inventories, or other data which
can be made available means withholding information important
to decision making.”

Tyler24 states that the most important principle in testing

 

23Association for Supervision and Curriculum Development, National

Education Association, Perceivigg, Behavi , Becomigg (washington, D.C.:
N. E. A., 1962), p.132.

24Leona E. Tyler, The Wbrk‘gg_the Counselor, Second Edition
(New York: Appleton-Century-Crofts, 1961), p. 106.

9

is that information obtained from guidance testing is for the use of
the client himself, and that our aim should be to enable the pupil to
form a sound idea of his own assets and liabilities. She feels that
it is better to do no testing at all rather than to Open the door for
the many kinds of misconceptions and anxieties that testing without
interpretation creates.

Dr. Dorothea MtCarthy of Fordham University seems to feel quite
strongly about the desirability of interpreting test results:

For a long time tests were administered in schools and the
practice was considered generally desirable, but the information
often remained in files on cumulative record cards and was not
used or interpreted to the testee himself or to his parents,
teachers, or counsellors. Such testing programs are utterly
useless and a waste of time and money unless the test results are
to be made known and unless they are to be used in decision.making
about the individuals who have been tested.25

Such comments are not new; one of the pioneers of psychological
testing, E. L. Thorndike, was saying about the same things in 1924:

A.highly valuable possibility of service for tests, it seems
to me, has been but little recognized. Too often tests are given,
the data are tabulated, conclusions drawn therefrom are utilized
by supervisors, and methods are revised by teachers because of them;
but the pupils who write the tests are not informed of any of the
results except in those rooms where unsatisfactory conditions have
brought about attempts to shift the blame to these pupils.

The final justification for every testing regime rests in
Nbry Jones and John Smith, and it therefore behooves all persons
who are making and giving tests to take them into partnership as
soon as and as completely as is feasible.26

 

25Dorothea HeCarthy, ”Ethical and Professional Considerations in
Reporting Test Information,". Proceedings and Summaries, ggg Annual
Meeting, leg York State Psychological Association, 11335., May 5-7, 1960,
pp. 11-12e

26E. L. Thorndike, “Tests and Their Uses," Teachers College
Record, XXVI, 2, October, 1924, pp. 93-94. .

10
VII. DISUSES AND MISUSES 0F STANDARDIZED TESTS

One of the fundamental goals of guidance is to assist the individual
student to understand himself, and results from standardized tests can
provide information that will aid in the attainment of this goal. In
order to fulfil this potential, however, tests must be chosen,
administered, scored, and interpreted correctly. Unfortunately, much
disturbing evidence is available to support the contention that far too
many tests are improperly handled during one or more of these steps.

Tor example, scoring for most standardized tests seems to be a
simple procedure. Yet Durost27 reports that from.101.to 502 of the answer
sheets coming in to his I. B. N. test scoring service have to be cleaned
up with erasers before accurate scoring is possible, and that 251 to 332
of teacher-scored tests coming in have to be conpletely rescored to provide
enough accuracy for mass statistical treatment. Phillips and Weathers28
tabulated the errors made by teachers in 5,017 scorings of parts of the
Stanford Achievement Test, and found that 282 of the tests contained one
or more errors in scoring. Test scoring is said to be the easiest and
most objective phase in the utilization of the instruments; if frequent

and serious errors are discovered at this stage, it is difficult to make

 

27H. N. Durost, "Present Progress and Needed Improvements in
School Evaluation Programs," Educational and Psychological Measurements,
14’ ppe 247-254e

28Beeman N. Phillips and Garrett Weathers, “Analysis of Errors
‘Nade in Scoring Standardized Tests," Educational and Psychological
Measurement, 18, 1958, pp. 563-567.

11
optimistic inferences about the more complex processes involved.
We have no reason to believe that every student who takes a
standardized test later receives an accurate interpretation of the
results; in fact, we cannot be sure that even a majority of them receive

the results in any form. Hastings and associates29

surveyed 1,000
eleventh‘grade students in 38 randomly-selected high schools in
Illinois, obtaining the following responses in percents to the question,

"Does the counselor discuss your test results with you?"

 

 

Curriculum Never Rarely Sometimes Frequently
College prep 15 14 42 29
Business-commercial 23 21 40 16
Vocational 26 21 38 15
Other 37 18 45 9

In the previously-mentioned survey by Gibson,30 his respondents revealed
interesting feelings toward test interpretation. When asked if they
understood what their test results mean, 43% answered “not sure" and
"no." Slightly more than 502 said they would have liked further
interpretations of their scores. As far as self-understanding is con-
cerned, only 44% indicated that they understood themselves to their

own satisfaction, and approximately the same percentage said they felt

their teachers seemed to understand them.

 

29Hastings, ggugl. "The Use of Test Results - Cooperative Research
Project Number 509" Bureau of Educational Research, University of
Illinois, 1960.

30Robert L. Gibson, "Pupil Opinions of High School Guidance Pro-
grams," [Personnel and Guidance Journal, XL, 5, January, 1962, pp. 453-457.

12
No interpretation may be preferable to improper interpretation.
Barry and Wolf state: "Perhaps the most damaging abuse of testing is

the incredibly poor test interpretation done in many schools."31 They

 

also take the position that nothing is more dangerous to the individual
than the mis-interpretation of test results.

Arnold32 developed a questionnaire about the Kuder Preference Record
and administered it to 200 entering college freshmen. He found that 44%
of them had taken the Kuder Preference Record while in high school.
Examination of the responses led Arnold to conclude that the written state-
ments of what these students said they learned from Kuder Preference Record
showed confusion of interests and abilities, and of measured interests with
other interests. He discovered that a "considerable number" of these
inventories had been given without the results being discussed later
with the students. Arnold speculates that inadequate test interpretation
in both quantity and quality must take the blame for this situation.

Berg33 tabulated the misunderstandings he noted in 30 clients
of six counselor trainees at Northwestern University. All of these
counselors held the M.A. in either psychology or education (guidance),

and all of them had some full-time counseling experience prior to entering

 

31Ruth Barry and Beverly Wolf, Epitaph for Vocational Guidance
(New York: Bureau of Publications, Teachers College, Columbia University,
1962), Pa 52a

32Dwight L. Arnold, "Student Reaction to the Kuder," Personnel
and Guidance Journal, September, 1958, pp. 40-42.

33Irwin A. Berg, "Test Score Interpretation and Client Confusions,"

Personnel and Guidance Journal, May 1956, pp. 576-578.

13
the practicum course. Types of test score misunderstandings revealed by

the clients and their frequencies of occurrence were as follows:

Gentile confusion with IQ 13.4%
Confusion over what IQ means 16.7%
Nbrm group confusion 40.0%
Confusion of interest and aptitude 50.0%
Scores a guarantee of success 56.7%

Berg feels that most of this confusion is the fault of erroneous counselor

conceptions and careless interpretations.

Finally, Leo Goldman in his book Usipg Tests ingounseling takes

 

the following position.
Finally, there is an alarming amount of misuse and disuse
of test results....From the vantage point of this observer,
the impression has been that many, if not most, of the tests

currently in use are either wasted or, even worse, used in
such a way as to misinform and mislead.34

VIII. STATEMENT OF THE PROBLEM

With the above factors in mind, this study was designed in order
to develop an approach to the communication and interpretation of test
scores, and to use the procedure with a group of seventh grade pupils
in order to investigate its effectiveness. The problem to be investigated
in this study is the effect of group interpretation of aptitude test scores
upon the self-estimates of abilities and the self-estimates of test results

of a sample of seventh grade pupils.

IX. SUMMARY

The extent of standardized testing has more than kept pace with

 

34Leo Goldman, Using Tests ip_Counseling, (New York: Appleton-
Century-Crofts, 1961), p. 2.

 

14
the growth of guidance and counseling practices in our schools.
Potentially, results from such instruments can provide students with
information valuable in the formation of accurate self-concepts, but
available evidence suggests that most of the results of such tests now
being administered are disused if not actually misused. The purpose
of this study was to develop a test interpretation procedure which would
increase the effectiveness of the interpretation process.

The following chapter presents a comprehensive review of the related
literature. Chapter III contains an account of the design and methodology
of the study, followed by a presentation of the findings in Chapter IV.

A summary of the study, the conclusions of the study, and some recommendations

for both practice and further research appear in Chapter V.

CHAPTER II
A REVIEW OF RELATED LITERATURE
I. RECENT RESEARCH

Until fairly recent times, directions for test interpretation
have explicitly or implicitly stated that it is to be done in the one-
to-one relationship of the counseling interview. Some writers in
the field still feel that this is the only acceptable approach. Barry
and Wblf, for example, typify this point of view:

Group interpretation of tests is usually even more disasterous
than the ordinary individual approach. The teacher or counselor
attempting the interpretation cannot take into account all the
feelings and values of thirty-odd students. In an attempt to
handle an impossible situation, he is likely to become either
so vague as to be unintelligible or so authoritarian as to eliminate
discussion. . . . Students are not, and should not be expected
to be experts in measurement. It is the counselor's responsibility
to help the students to develop some comprehension of what their
own scores mean.1

Nevertheless, in the past few years interpretation of test
results to groups of students seems to have become more and more common.
Perhaps the original reason.was an attempt to utilize valuable and
limited counselor time more efficiently; any necessary compromise in
effectiveness was thought to be a better alternative than no inter-
pretation at all. But some workers in the field, probably beginning
with Proehlich, began to suggest that not only was group interpretation

just as effective as individual interpretation in many cases, but

perhaps even more so. A search through the literature of the past

 

1Ruth Barry and Beverly Wolf, Epitaph for Vbcational Guidance,
(New York: Bureau of Publications, Teachers College, Columbia University,
1962), pp. 53-54.
15

l6
decade does not produce a great number of studies of group interpretation,
but it must be added that a similar search would not disclose many
studies dealing with ppy_type of test interpretation. The following
seem to be the major ones that have produced implications for this
particular study.

Froehlich and Mosher2 studied what they considered to be a
logically expected outcome of counseling, the memory of test scores.
They did group and individual interpretations of Differential Aptitude
13253 scores with 150 ninth graders, which include having them draw
their own profiles which they kept. Fifteen months later, the ex-
perimenters found that a large proportion of the students did not
report accurately when they were asked to redraw their D. A. T.
profiles. Although high scores were remembered more accurately than
low ones, there was a tendency for both high and low scores to be
reported as being closer to the mean than they actually had been. The
authors contend that this finding contradicts the "general belief" among
counselors that it is easier for high rather than low ranking pupils
to understand and accept test evidence, if one assumes that accuracy
of self-report fifteen months later is closely related to understanding
and acceptance.

Two evaluative comments seem applicable. First, the unequal

size characteristic of percentiles may have functioned to make memory

 

2C. P. Froehlich and W. E. Maser, “Do Counselees Remember Test
Scores?" Journal p§_Counseling Psychology, 1, Fall, 1954, pp. 149-
152.

 

17
appear worse than it actually was. For example, is it fair to say that
the same degree of memory inaccuracy is demonstrated by the student who
remembers his score as the 80th percentile when it acutally was the
90th, and the student who remembers his score as the 50th percentile
when it actually was the 60th? Secondly, the fifteen-month time
interval actually represented a period starting early in the freshman
year in high school to about a third of the way through the sophomore
year. During this period of time the students might have grown in
maturity, competency, achievement, and self-understanding to the point
that the profile drawn then was the one they felt applied to them at
that time, and not a reproduction of something that might have been
true of them over a year previously.
Lane3 did individual interpretations with 111 high school students,
comparing the effectiveness of what he calléd the traditional, directive,
counselor-centered method with that of a permissive, non-directive, client-
centered method. He found no significant differences. Two criterion
measures of self-understanding were used, a check list and an essay, but
the experimenter reported disappointment at the low level of performance
on the essays. Low positive relations were found between check list
scores and measures of scholastic aptitude, socio-economic background,

and even less for social adjustment, although all were declared “minor

influences."

 

3David Lane, “A Comparison of Two Techniques of Interpreting
Test Results to Clients in Vocational Counseling," (unpublished
doctoral dissertation, Columbia University, 1952).

18

Lallas4 compared three methods of interpreting achievement test
scores: (1) individual counseling interviews, (2) group interpretation,
and (3) a combination of the individual plus group approaches. His high
school junior subjects estimated their rankings on the various subtests
of the I232 233 £19; Educational Development before and after inter-
pretation of their own scores. All three experimental groups showed
significantly greater improvement in accuracy of self-estimate than the
control group, with greatest improvements appearing in the individual
counseling and the group-plus-individual counseling groups, with somewhat
less improvement shown by those receiving only group interpretations.

Generalization and application of the results of this experiment
are limited by weaknesses in the design and procedure. The experimenter
attempted to match subjects in his experimental groups; the group-plus-
individual counseling experimental group received much more attention than
the others, and seven different graduate student counselors did the individual
interviews while one individual did all of the group interpretations.
Specific details of what actually transpired in the group interpretation
sessions is missing, and only vague details of the individual interview
procedure is given.

Wright,5 working with college students, found that while both

individual and multiple or group interpretations resulted in significantly

 

“John E. Lallas, "A Comparison of Three Methods of Interpretation
of Achievement Tests to Pupils," (unpublished doctoral dissertation,
Sanford University, 1956).

5Edward Wayne Wright, "A Comparison of Individual and MMltiple
Counseling in the Dissemination and Interpretation of Test Data," (un-
published doctoral dissertation, University of California, 1957).

19
more accurate self-reports than no interpretation at all, he could
demonstrate no meaningful differences between the individual and group
approaches. Unfortunately, details of the procedures used in either
method are not given.

Froehlich,6 in one of the rare published studies that utilized
high school students, concluded that his data did not support the
claim that counseling (which in this case involved the interpretation of
the Differential Aptitude Tests) must be individual. Using several
methods of data analysis, he found no significant differences between
individual and group interpretation in all methods but one, and in that
instance the group approach was significantly more effective. His
design seemed to have several weaknesses, however, including non-
randomization of volunteer subjects, varying group size, and only a total
of 42 subjects involved. No details or descriptions of either the
individual or group interpretations are available.

Gustad and Tuma,7 using both client-centered and more directive
techniques of test score interpretation with male undergraduate college
students, found no significant differences in client learning about self.
Moreover, they could demonstrate no significant relationship between

scholastic aptitude and client learning about self during counseling.

 

6Clifford P. Froehlich, “Must Counseling Be Individual?"
Educational ppg_Psychological Measurement, XVIII, 4, 1958, pp. 681-689.

7John w. Gustad and Abdul a. Tuma, "The Effects of Different
Methods of Test Introduction and Interpretation on Client Learning in
Counseling," Journal 2; Counselipg Psycholpgy, Vol. 4, No. 4, 1957,
pp. 313-317.

20
They did find that initial accuracy of self-ratings was positively related
to client learning, suggesting that those who knew themselves best before
counseling learned the most about themselves during counseling, assuming
that the test scores involved represent the "truth" about their interests
and abilities.

L. B. Rogers8 found that a self-evaluative technique in which non-
test data were surveyed before test results were introduced and in which
client participation was encouraged was no more effective than a test
centered method in which the profile was explained in detail but no
particular effort was made to stimulate client participation or to in-
troduce non-test data. His college student subjects given the one type
of interview showed no greater growth in self-understanding than those
given the other kind, when both groups included a wide range of intelligence
and of responsiveness in the interview.

Some modest support for the value of greater client participation
was found when each of the two main treatment groups was subdivided into
two groups, those who had participated actively in the interview and those
who had not. The sub-group which participated actively in the self evaluative
interview showed a significant increase in self-knowledge; others increased
insignificantly.

The factor of intelligence as measured by the A_._ _C_._ E; Psychological
Examination was also studied. When the highest and lowest four deciles

in each of the two main treatment groups were studied, it was found that

 

8Lyle B. Rogers, "A Comparison of Two Kinds of Test Interpretation
Interview," Journal pf Counseling Psychology, 1954, 1, pp. 224-231.

21
the self-evaluative approach was as effective with the less intelligent
as with the more intelligent, whereas the test-centered approach was not
as effective with the less intelligent as it was with the more intel-
ligent.

Evidence concerning the role of the counselee in test interpretation
is provided by a study conducted by Dressel and Matteson at Michigan
State? Forty recorded interviews with freshmen college students with
seven counselors were used. They tested the hypotheses that in comparison
with counselees who participate less in the test interpretation process,
those who participate more actively (1) gain more in self-understanding,
(2) are more certain of their vocational goals, and (3) are more satisfied
with the experience. Their findings were moderately supportive of the
first and second hypotheses, but not of the third.

Greater gains in self-understanding were made by counselees whose
counselors succeeded in eliciting the greatest amount of counselee
participation in the interview, and they were more certain of their
vocational goals afterward. However, their results do not indicate
that clients who participate more get more out of the experience regardless
of who the counselor is and what he does.

Of particular importance to this study are the findings of an

investigation conducted by Kamm and Wrenn at‘Minnesota.10 They set out

 

9Paul L. Dressel and Ross W. Matteson, “The Effect of Client
Participation in Test Interpretation," Educational and Psychological
‘Measurement, X, 4, Winter, 1950, pp. 693-706.

10Robert B. Xamm and Gilbert C. Wrenn, "Client Acceptance of
Self-Information in Counseling," Educational and Psycholpgical
Measurement, X, 1, Spring, 1950, pp. 32-42.

22
to study within which interview situations counselees do and do not tend
to accept presented information, how those who do accept information
differ from those who do not, and what type of information does and does
not tend to be accepted. "Acceptance“ was defined as fevorable reception
by the client of information presented to him, as demonstrated by what
the counselee said and did. "Information" was defined as all data presented
by the counselor, whether they be in the form of advice, suggestions,
emphases, recommendations, interpretations, requests, or explanations.
Forty recorded educational-vocational planning interviews with one trained,
experienced counselor were analyzed, with additional data coming from
post-interview check lists and follow-up interviews one and four months
later.

They found that both acceptance and non-acceptance of information
occured in situations in which the counselee-counselor relationship was
judged to be friendly, suggesting that evaluation of effectiveness cannot
be done validly by measuring good will.

With the exception of information involving alternation of previously-
made counselee plans which tended to be more often accepted by the acceptance
group, different kinds of information were accepted equally well by acceptance
and non-acceptance groups.

No relationship was found between acceptance of information and
any of the following factors: academic aptitude, particular measured
personality patterns, social class, veteral status, marital status, previous
counseling interviews, length of interview, or proportion of time which

the client spoke during the interview.

23

They found that information is likely to be accepted if it is
presented in an emotionally-relaxed atmosphere; is directly related to
the counselee's own immediate problem and felt needs; and is congruent
with and not in opposition to his self-concept. The authors state that
perhaps the most conclusive of all their findings is that the client
himself is the basic determiner of whether or not acceptance occurs. They
suggest that the feelings, needs, wants, desires, and attitudes of the
counselee are more important than the characteristics of the interview
situation or type of information presented.

The authors recommend that counselors use techniques designed to
assist in the development of the counselee toward a more realistic aware-
ness of himself, and that they be aware of the level of thinking of their
counselees.

Singer and Stefflre11 investigated the effects of counseling upon
the expressed self-interests of high school seniors, analyzing both means
and standard deviations for significance. For the boys, they found no
significant change in means, and for the girls, significant mean changes
only for the Science and Mechanical scales. When standard deviations
were compared, it was found that all of the standard deviations for the
boys decreased, although only for the Mechanical scale was this decrease
significant. For the girls, four of the six standard deviations decreased,

with significant decreases on the Science and Mechanical scales. Thus

 

ll'Stanley L. Singer and Buford Stefflre. "Analysis of the Self-
Estimate in the Evaluation of Counseling," Journal pf_Counseling_Psychology,
I, No. 4, Winter, 1954, pp. 252-255.

24
only five of a total possible twenty-four comparisons were statistically

significant.
II. LIMITATIONS OF PREVIOUS STUDIES

The following are limitations of the studies of test interpretation
which have been reported:

1. In most cases only a small number of subjects was used.

2. In most cases sampling methods violated the most fundamental
rules. More than one utilized volunteers, and randomization of subjects
to various treatments was seldom.mentioned.

3. Group size varied greatly, sometimes within the same study.
Seldom was group size near what is considered to be a typical classroom
group. Few groups in available studies approach as many as a dozen
subjects at a time.

4. An examination of the original sources produces few details
of what was said and done in the interpretation sessions. Specific
replications would be most difficult if not impossible.

5. Most of the studies were performed at the college level, only
a few used high school students, and none used students at the junior
high school level. If concept of self is involved in the communication
of test results, it seems that investigations should be undertaken at

that stage where the self-concept is being formed.
III. SUMMARY

To summarize, available research pertinent to this problem seems

to justify the following generalizations:

25

1. No consistent superiority of any one particular approach
(ex., directive vs. non-directive) has been demonstrated.

2. Techniques of group interpretation have not been clearly
demonstrated to be inferior to techniques using individual interpretation.

3. In every available case, individuals who have had their test
scores interpreted later demonstrate a higher degree of accurate self-
knowledge than those who received no interpretation.

4. Acceptance of test results has not been shown to be dependent
upon level of mental ability as measured by traditional standardized
academic aptitude tests, measured personality patterns, social class,
level of performance, or counselee experience in the counseling setting.

5. Acceptance of test results appears to be related to the status
of the counselee's self-concept as it is either threatened or strengthened
by the information being presented.

6. Counselors who succeed in eliciting counselee participation
in the interpretation process are likely to make possible greater gains
in self-understanding by the counselees.

7. Although moderate successes have been demonstrated in communicating
the results of testing, a distressingly large proportion of counselees
seem to come away from interpretation sessions little if any wiser about

their characteristics than before.

 

CHAPTER III

DESIGN AND METHODOLOGY OF THE STUDY

I. THE PROBLEM

The results of research studies reported in the previous chapter
suggest that techniques of group test interpretation can be improved,
and also suggest the directions in which further investigations should
proceed.

This study was a combination of the experimental and descriptive
approaches. Essentially, the purpose of the study was twofold: (l) the
development of an approach to the communication and interpretation of the
concepts of individual differences, norm groups, and the meanings of test
results, and (2) the use of this approach with a large group of seventh
grade students in order to gage its effectiveness.

Following procedures reported in previous studies in this area,
interpretations were made to groups of subjects who estimated their test
results prior to and after this interpretation. The interpreter attempted
to remain neutral toward the test data and permissive toward the students'
reactions to these data. Pupils were invited to request individual counseling
interviews in case they wanted to discuss their results either further or
privately.

This study goes beyond previous ones in several ways. It involved
students in the seventh grade, an earlier grade level than is reported
in the literature. A larger number of subjects was involved. In addition

to estimating test results, subjects also estimated their abilities

27
in those areas sampled by the test. Estimates were based upon a stanine
scale, which utilized a combination of numbers and cartoon symbols.
(See Appendix B, p.114.) It was felt that this device had many advantages
over the traditional thermometer-like percentile graph in communicating
the concepts of individual differences and of norm groups. The one hundred
cartoon symbols were used in an attempt to visually sumbolize these
concepts. Estimates in the form of stanine bands rather than in percentiles
were used in order to lessen the temptation to think of a score as a point
rather than as a general area, and to eliminate the disadvantage of the
unequal units of the percentile distribution. Moreover, stanines can be
processed mathematically without having to be transformed statistically.

As another innovation, the presentation of test results was preceded
by a presentation and discussion of the concepts of individual differences
and of norm or reference groups. It was reasoned that if the pupil had
some prior understanding of the natural and expected variability in people,
the theory of the normal curve of distribution, and the nature of norm
groups, he would be able to accept the test results which might otherwise
be rejected.

This experimental approach was designed to involve the individual
pupil at every step. The pupils estimated on a mimeographed form their
approximate positions in a normal stanine distribution of their peers.

(See Appendix C, p.1J13) Test results introduced as the test's estimate
of these same approximate positions, may or may not have agreed with the
pupil's estimates. In either case, he was free to either alter his thinking

about himself or to reject the test results. No value judgment was placed

28
upon the degree of agreement with test results. That is, the individual
who disagreed with the implications of the test results and who indicated
this was not judged to be unrealistic in his opinions. Obviously, the
test results may have been, for him, somewhat inaccurate. In other words,
the assumption was made that test scores can be wrong.

The exPerimental estimation procedure and interpretation procedure
were gleaned from recommendations of authorities in the fields of guidance
and testing, from those descriptions that can be found in the original
sources of previous studies, and from the implicit and explicit recom-
mendations of the test publisher. It was felt by the eXperimenter that
these methods are consistent with good practice as it is known and accepted
at this time. Complete details of what was said and what occurred during
the two meetings with each experimental group can be found in Appendix A,

page 104, and Appendix D, page 118.

II. LIMITATIONS OF THE STUDY

Several limitations are apparent in this study that have an effect
upon the conclusions which can be drawn from the data. The following are
recognized shortcomings:

1. Only one week intervened between the date of interpretation
and the follow-up collection of data. This period of time was chosen
partly by the realities of the situation. However, it is difficult to

accurately measure the effect of time on the ability of the individual

29
subjects to recall information obtained in test interpretation inter-
views.

2. The experimenter did all of the estimation and interpretation
presentations. Therefore the results may have been at least in part
a reaction to his personality and techniques.

3. The sample consisted of seventh graders only, selected
from a single school. The extent to which the findings of the study
can be generalized are therefore limited to the extent to which this
sample represents seventh graders in general, and to the extent that
the performance of seventh graders is like that of pupils at other
levels.

4. Rapport between the investigator and the various groups was
an unmeasured factor. While the impression of the experimenter was that
rapport was successfully established between himself and the various
groups, no objective evidence is available to either support or refute

this opinion.

III. SAMPLE SELECTION

Two major factors influenced the decision to conduct this study at
the seventh grade level. First, the seventh grade typically coincides
with the beginning of adolescence, a period in which a concept of self
is said to be emerging as onecf the products of eXperience. The individual's
conceptions of his abilities are an integral part of this self concept.

Second, as mentioned previously in Chapter II, most of the published

30

investigations of test interpretation have been conducted with college
students, a few with high school pupils, but none at the junior high school
level. For these reasons it seemed apprOpriate to use seventh grade boys
and girls as subjects for this study.

All subjects were drawn from the seventh grade class of one junior
high school, WOodrow Wilson Junior High School, Terre Haute, Indiana.
This school offered a large number of subjects drawn from at least three
readily identifiable socio-economic areas of the city. The principal
and the faculty were receptive to assisting with carrying out this inves-
tigation. Although some guidance services were being provided, because
of a lack of counselor time, test interpretation to the pupils was not
at that time being done. All students in the seventh grade class were
tested. Only those students present on the dateof the test, the day of
the interpretation, and the day of the follow-up were included in the
study. Table I on page 31 summarizes the number of subjects involved
in the study. Ten groups of seventh grade pupils averaging 28.1 pupils
per group received the interpretations. Complete data were collected from

244 subjects, 119 girls and 125 boys.

31

 

 

¢¢N mHH mNH HwN NMH ¢¢H Huuoa

 

 

Hm NH mH qm mH mH oH<m
aN NH NH Hm mH NH ¢¢s
¢N m 0H 5N 0H NH w<~
MN HH NH on nH NH m<~
nN HH «H mN NH 0H 045
«H n s 5H m CH m<m
NN NH oH 5N mH «H «<5
NN oH NH mN ¢H mH m<~
Hm 0H mH mm 0H NH N<~
MN mH w mN oH m Hdn
HouOH mHuHo whom Huuos mHuHo uNom umnaaz adouu
.umu mmMoucH umom dawns coHuwuouauoucH adouw‘GOHusuounuoucw sH nonadz

noun ..H.m.<uumuoofin9o «Home: mo nonssz

 

 

NQDHm mma 2H nm>AO>ZH maomnmam mo ZOHHbmHmHmHn

H mqm<a

32
IV. SELECTION OF THE TEST

The test chosen for administration and interpretation was the
Academic Promise Tests, an aptitude battery recently published by the
Psychological Corporation of New York. In part an outgrowth of the
Differential Aptitude Tests, the A.P.T. battery was designed for use in
grades six through nine. A brief description of these tests follows; a

more complete description appears in the A.P.T. Manual.1

 

Four types of tests compose the battery, each selected with certain
principles and objectives in mind. The verbal section was designed to
measure the understanding of word meanings and the ability to use words
in reasoning. Items are of the analogies type. The numerical section was
designed to measure the capacity to think in quantitative terms, to under-
stand and use numerical relationships. Items are of several types, emphasizing
the understanding of concepts and reasoning. The reading requirement for
this section is low as few words are used. The abstract reasoning section
was designed to measure the ability to see relationships and to recognize
concepts presented in the form of diagrams or symbols rather than in words
or numbers. Items are of the figure classification type, requiring the pupil
to seek out the principle which provides a common characteristic for a set
of three figures, and to recognize which of several other figures shares

that characteristic. The language usage section was designed to measure

1George K. Bennett, Marjorie K. Bennett, Dorothy M. Clendenen,
Jerome E. Doppelt, James H. Ricks, Jr., Harold C. Seashore, and Alexander G.
Wesman, Academic Promise Test Manual, 1962, New York: The Psychological
Corporation, 1962.

33
understanding of correct writing and speech, together with the appreciation
of good English. Items require the pupil to identify errors in grammar,
spelling, and punctuation. Items on all sections are arranged in order
of increasing difficulty. Total testing time is ninety minutes.

A common practice in measurement is to group abilities into verbal
and non-verbal classifications. With the A.P.T., the verbal and language
usage tests can be combined to provide a measure of broad competence with
verbal materials. In a similar fashion the numerical and abstract reasoning
sections together provide a measure of broad competence with non-verbal
materials. Finally, all four scores when added together provide a single
statement of general academic aptitude.

Norms for the A.P.T. are based upon the performances of more than
34,141 pupils in grades 7-9, including 9,141 seventh graders. This norms
group was put together by a careful stratified sampling procedure which took
into account pupils enrolled by grade level, geographic region of the nation,
and community size.

A summary of validity data is given in Table II on page 34 and

a summary of reliability data appears in Table III on page 35.

34

TABLE II

MEDIAN VALIDITY COEFFICIENTS BETWEEN A.P.T. SCORES AND SCHOOL GRADES2

 

 

 

AR N V LU AR+N VfLU A.P.T. Tot.
English .31 .50 .45 .60 .45 .59 .58
Mathematics .30 .58 .36 .41 .54 .46 .56
Social Studies .29 .45 .41 .45 .40 .51 .54
Science .33 .49 .46 .46 .48 .50 .56

 

 

 

2George K. Bennett, Marjorie K. Bennett, Dorothy M; Clendenen,
Jerome E. Doppelt, James H. Ricks, Jr., Harold C. Seashore, and Alexander G.
WEBman, Academic Promise Test Manual, 1962, Tables 4, 5, 6, and 7.
New'York: The Psychological Corporation, 1962.

 

35

 

 

TABLE III
RELIABILITY DATA FOR THE ACADEMIC PROMISE TESTS3
GRADE 7
=— _=
A.P.T.
AR N V LU AR+N VdLU Total
Alternative-form reli-
bility coefficients .82 .87 .82 .88 .88 .90 .93
Standard error of
measurement in points
of raw score (maxi-
mum score on each of
the first four sect-
ions - 60) 5.1 3.4 3.6 3.7 6.7 5.6 8.7

__‘

3George K. Bennett, Marjorie K. Bennett, Dorothy M; Clendenen,
Jerome E. Doppelt, James H. Ricks, Jr., Harold C. Seashore, and Alexander G.
Wesman, Academic Promise Tests Manual, 1962, Tables 23 and 24. New'York:
The Phychological Corporation, 1962.

36

It was decided to report to the pupils only the verbal, numerical,
language usage, and abstract reasoning scores, and to ignore for the purposes
of this study any consideration of either the verbal, non-verbal, or total
A.P.T. scores. The reasons for this decision were as follows:

1. While the four basic abilities measured are relatively concrete,
objective, and easy to eXplain and understand, the combinations of them
are more abstract, subjective, and difficult to explain and understand.

2. The total A.P.T. score can be interpreted in much the same manner
as the score from "mental ability" or "group intelligence" tests. It was
felt that the introduction of such concepts might introduce unwanted and
contaminating emotional factors into the learning situation.

3. It seemed that four factors were enough to deal with in an
introductory investigation of test score interpretation at this grade

level.
V. COLLECTION OF THE DATA

On Monday, February 4, 1963, all seventh grade pupils of WOodrow
Wilson Junior High School, Terre Haute, Indiana, took Form.A of the

Academic Promise Tests. These tests were administered in the pupils'

 

first period classrooms by their regular first period teacher and a college
student assistant trained by the experimenter in group testing procedures.
During the next month, answer sheets were scored, individual profiles were
prepared, and group statistical analysis was begun.

On Menday, March 4, 1963, and Tuesday, March 5, 1963, the experimenter

returned to the school for the second phase of the study. This step

 

1: VI. I III! I [III III I
«I'll 'Illl II III I.

37
required one class period per group. A complete record of what was said
and done appears in Appendix A, pages:104 to 112. Using a chart illus-
trating a norm group (Appendix B, page 114), and an estimation worksheet
(Appendix C, page 116), each pupil made an estimate of his or her abstract
reasoning ability, numerical ability, verbal ability, and language usage
ability. Also each pupil estimated his or her test score in these areas.
The two sheets were then collected. The final step of this second phase
of the investigation was to interpret the test results, using the test
publisher's profile sheets which had been drawn previously for each pupil.
(A complete record of this interpretation presentation appears in Appendix
A, pages 104 to 112).

Exactly one week later, the experimenter returned to each group
for the third phase of the investigation. Following the procedure
presented in Appendix D, pages 118 to 123 , and using the norm group chart
(Appendix B, page 114), and another estimation worksheet (Appendix C,
page 116), the pupils re-estimated their test scores and abilities in the
four areas sampled by the test.

Thus each student contributed the following data: before he received
interpretations of his test scores, stanine estimates of his abstract
reasoning ability, numerical ability, verbal ability, and language usage
ability, together with an estimate of his test scores in each of these
areas, plus the same estimates one week after his test scores had been
interpreted. Since estimates were on a stanine scale, four numbers made
up each of the two preinterpretation estimates, and four composed each of

the two postinterpretation estimates.

38

VI. ANALYSIS OF THE DATA

Computation g; the estimation scores. Raw scores for each of the

 

sub-tests of the A.P.T. were converted first to percentiles (based upon
the national norms of the test publisher), and then to stanines. The
discrepancy of each estimate stanine from the actual test score stanine
was computed, and the four numbers for each of the four estimates were

summed algebraically. The following hypothetical data will serve as an

 

 

 

 

example.
Preinterpretation Postinterpretation
Estimates Estimates
AR N LU V AR N LU V
Test Scores 9 7 5 4 9 7 5 4
Abilities
Estimates .2 _s _5. .2 _§ __6; .2 .2
-4 -2 0 +1 -1 -l 0 +1
3 -5 a —1
Test Scores 9 7 5 4 9 7 5 4
Test Results
Estimates __6 __6. __5_ _5 9 __8 _5 _4
'-3 -1 0 +1 0 +1 0 0
a -3 =+1

 

In order to avoid minus numbers, a constant of +32 was added. This
sum was selected because the maximum.under-estimation possible is -32.
Such an event could occur if an individual had four stanine test scores
of 9, but made four estimates of 1, thus scoring 4(-8) on the scale.

Likewise, the maximum over-estimation would be 32, which could occur if

39
an individual had four stanine test scores of l, but made four estimates
of 9, thus scoring 4(+8) on the scale. With the constant added, the
final scale runs from 0 to 64, with 32 marking the point of no discrepancy
of estimates from test scores. Scores above 32 indicate over-estimations,
scores below 32, under-estimations.

In the example above, adding the constant of +32 to each of the four
preliminary estimation sums produces the following estimation scores:
preinterpretation estimate of abilities: 27; preinterpretation estimate
of test results: 29; postinterpretation estimate of abilities: 31;
postinterpretation estimate of test results: 33. All data for an
individual were in this manner reduced to four numbers, one for his pre-
interpretation estimate of his abilities, one for his preinterpretation
estimate of his test scores, one for his postinterpretation estimate of
his abilities, and one for his postinterpretation estimate of his test
scores. These are the data that were analyzed statistically for significance
of change.

An arbitrary decision was made by the experimenter not to investigate
in this study estimation changes for the various individual abilities and
individual tests of the A.P.T., or to compare the accuracy of estimation
of various combinations of these individual abilities and individual tests.
That is, it was decided not to pursue such questions as how the pupils
performed in estimating their verbal ability or if there was any meaningful
difference in their accuracy of estimating verbal test scores as compared
to abstract reasoning test scores. The rationale for this decision is

as follows.

 

lII|II.I..|| lllllll'llllu

 

40

First, it should be noted that the four abilities under discussion
here are not completely separate and independent mental factors. Neither
are the various tests of the A.P.T. completely separate and independent.
Table IV on page 41 presents the mean intercorrelation coefficients of
the A.P.T. Examination of these data reveals considerable relationship
between the individual tests and combinations of tests.

A second factor somewhat related to the first is the relationship
of the length of a measuring instrument or procedure to its reliability.
It seemed reasonable to assume that a total estimation score formed by
summing four abilities which are to some degree related would be more
reliable than an estimate of one fourth that size, the estimate of a single
ability.

Finally, it was felt that these topics were not the main focus of
an investigation which was to be one of the first studies of test score

interpretation at this grade level.

41

TABLE IV

MEAN INTERCORRELATION COEFFICIENTS OF THE A.P.T.4

 

 

 

N V LU AR+N V+LU
Abstract Reasoning .55 .51 .37 -- .49
Numerical .69 .59 -- .70
Verbal .63 .67 --
Language Usage .54 --
AR + N -- .66

 

 

 

4George K. Bennett, Marjorie G. Bennett, Dorothy M. Clendenen,
Jerome E. Doppelt, James H. Ricks, Jr., Harold C. Seashore, and Alexander G.
wesman, Academic Promise Tests Manual, 1962. Table 26. New York: The
Psychological Corporation, 1962.

 

42
Data analysis procedures. Means, variances, and standard deviations
of the preinterpretation and postinterpretation estimates of abilities
and of the preinterpretation and postinterpretation estimates of test
results were computed for the following groups: the total study group;
boys; girls; the overestimators of abilities; the underestimators of
abilities; the overestimators of test results; and the underestimators

of test results.

Statistical techniques. Two statistical techniques were utilized

 

in this study, the t-test of significance of difference between means
and the F-test of homogeneity of variances.

This investigation focused on the effects of the interpretation of
aptitude test results upon the estimates of abilities and the estimates
of test results of seventh grade pupils. By means of the t-test, means
of the preinterpretation estimates of abilities of the total group and
of the various subgroups were compared with the postinterpretation estimates
of abilities of the same groups. In a similar manner, the means of the
preinterpretation estimates of test results of the total group and of
the various subgroups were compared with the postinterpretation estimates
of test results of the same groups. This was done in order to assess
the probability of significant change in mean estimates.

The F-distribution tests the hypothesis that two normal distributions
have equal variances. It is an appropriate technique in this situation
because an examination of means only may fail to identify important
differences in variability. In this study, for example, one individual

or two different individuals might earn estimation scores of 32 in quite

43
different ways. One such score could be earned by estimating all four
test scores accurately. However, it is also possible to earn an
estimation score of 32 by balancing two overestimations with two under-
estimations. Since estimations can have the similar means but quite
dissimilar variances, it is necessary to investigate possible significant
changes in variances of estimations as well as possible significant
changes in estimation means.

In this instance the F-test was used to ascertain the probability
of as great a difference (irrespective of direction, i.e., a hypothesis
or decision requiring a two-tailed test) between the variances for two
groups. 'McNemar5 says that where the significance of variability of two
groups is being tested, an F at the .01 point of the table means significance
at the .02 level; an F at the .05 level means significance at the .10

level; and an F at the .001 level indicates significance at the .002 level.

Level gf significance selected. For this particular study it was

 

felt that establishing a minimum level of significance less than .05

was not warranted. The .05 level of significance, therefore, was selected
as the required acceptable level of significance for this study. Any
difference so large that it would be expected to occur by chance alone
only 5 or less times in 100 will be accepted as evidence sufficient to

reject the hypothesis of no difference.

 

5Quinn McNemar, Psychological Statistics, (New York: John
Wiley & Sons, Inc., 1955), p. 246.

44

VI. HYPOTHESES TO BE TESTED

1. Postinterpretation estimates of abilities will not differ
in accuracy from preinterpretation estimates of abilities.

2. Postinterpretation estimates of test results will not differ
in accuracy from preinterpretation estimates of test results.

3. Preinterpretation estimates of abilities will not differ in
accuracy from preinterpretation estimates of test results.

4. Postinterpretation estimates of abilities will not differ
in accuracy from postinterpretation estimates of test results.

5. No difference in accuracy of estimates of abilities or of
estimates of test results will be found between the sexes, between the
high and low academic aptitude groups, between the over-estimators of
abilities and under-estimators of abilities groups, or between the over-

estimators of test results and the under-estimators of test results group.

VIII. SUMMARY

The Academic Promise Tests were administered to all seventh grade

 

pupils of one Indiana junior high school. Just before and one week after
group test results interpretation, the pupils estimated their abilities
in those areas sampled by the tests and their test scores. Means

and variances of the estimates were tested for significance of change
using appropriate statistical techniques. Null hypotheses predicted

no significant differences for the various combinations of estimates

45
for the total group and several subgroups. The results of these

statistical procedures appear in Chapter IV.

CHAPTER IV

FINDINGS OF THE STUDY

On the following pages of this chapter, the various statistical
findings of the study are presented, analyzed, and summarized.

After taking the four sections of the Academic Promise Tests,

 

each subject in the study contributed four scores: a preinterpretation
estimate of his abilities; a preinterpretation estimate of his test

scores; a postinterpretation estimate of his abilities, and a postinter-
pretation estimate of his test scores. Estimates were made in terms of
stanines, and in order to eliminate negative numbers, a constant of +32
was added. If, for example, an individual estimated all four of his

test scores in the same stanine band as the scores actually fell, his
estimate score was 32.00. Scores larger than 32 indicate over-estimations;

scores smaller than 32, under-estimations.

I. STUDY GROUP NORMS COMPARED TO NATIONAL NORMS

Means and standard deviations of test scores for both the study
group and the publisher's national norm group are presented in Table V,
on page 47. The means and standard deviations of the study group were
found to be very similar to those of the national norm group furnished

by the test publisher.

II. MEANS OF ESTIMATES

Means of estimates appear in Table VI, page 48. These means were

46

47

TABLE V

ACADEMIC PROMISE TESTS NATIONAL NORMS COMPARED TO STUDY GROUP NORMS

 

 

 

Means Standard Deviations

Norm Study Norm Study

Group Group Group Grggp
Abstract Reasoning 28.7 31.27 12.2 13.3
Numerical 25.6 25.76 10.0 9.8
Verbal 29.0 33.36 9.5 9.3
Language Usage 27.8 28.45 11.0 9.9
A.P.T. Total 111.1 115.5 36.1 34.9
Boys -- 115.97 -- 35.6

Girls -- 115.05 -- 34.1

 

 

48

TABLE VI

MEANS OF ESTIMATES OF ABILITIES AND POSITIONS OF TEST RESULTS
FOR BOYS, GIRLS, AND TOTAL GROUP

 

 

 

Boys (125) Girls (119) Total (244)
Pre- Post- Pre- Post- Pre- Post-
Estimate of
abilities 33.94 34.45 35.40 35.31 34.65 34.87

Estimate of
position of
test results 33.56 33.37 34.36 33.79 33.95 33.57

 

49
tested for significance of difference, the results appearing in Table
VII on page 50. No significant differences were found. This was not
an unexpected finding, since test of significance of difference of means
is not the most appropriate technique for this particular situation.
An individual with two extreme over-estimates and two extreme under-
estimates might have a total mean estimate not unlike the individual
who estimated all four parts accurately. It seems more logical to look

for changes in variability than in means.

III. VARIABILITY OF ESTIMATES FROM GRAND MEANS OF ESTIMATES

Table VIII, page 51, indicates the variances and standard deviations
of the estimates of the study group, while Table IX on page 52 shows
the results of testing the significance of variance of various combinations
of preinterpretation and postinterpretation estimates by use of the F-
test. After interpretation of test results, the total study group had
less variance in both the estimates of abilities and the estimate of test
results, at a highly significant level of confidence. For the total
study group, no significant difference was found between the variances
of the preinterpretation estimates of ability and the preinterpretation
estimates of test results. After interpretation of test results, the
total study group had less variance in the estimate of test results than
in the estimate of abilities at a highly significant level of confidence.
Table X on page 53 indicates the variability of the boys' pre-
interpretation and postinterpretation estimates of abilities and of test

results as computed from the means of the total group, while Table XI on

 

 

50

 

.. Ha.o- mo. mm. mn.nm om.¢m manmoM mama mo
moumawumm mo unmoz_.mHuHo

u- mH.o+ mo. mo. Hm.mm o¢.nm aoHuHHHc< «0
moumEHumm no mass: .mHuHu

.. om.o+ Hm. mH. mm.mm om.mm muHsmsm uan we
moumBHumm mo annex .mhom

.. mw.o- ow. Hm. m¢.¢m em.mn moHUHHHn< mo
mmumEHumm «0 name: .nmom

u- Hm.¢+ Ne. mm. mm.mm mm.mm manaom umma «0
moumEHumm mo «new: undue

.. m¢.o- we. NN. Nm.¢m mo.¢m moHuHHHnw mo
moumaHumm we anus: undue
magnifies; u mm a a8: :8 :8: 9.5

 

 

macaw A<HOH Mm 92¢ Nam um

mHHDmmM BmmH mo mme<2HHmm onH<Hmmmmm92HHmom 92¢.ZOHH<HMMNMMHZHMMA mo mz<mz ho
Qz< mmHHHHHm< ho mmH<ZHHmm ZOHH<HMMMMNHZHHmom nz<_ZOHH<HmmmmMHszmm ho mz<mz ho ZOmHM<mZDU

HH> MHQ<H

51

 

 

 

 

 

HowH.m NHmH.oH om.o~¢N Nm.¢oMH w¢.HHHH NH.MN+ mN.mNu manmmh umou «0
sumaaumouumom

onm.m oHem.mH oo.mwmm qo.moHN om.HmNH n¢.om+ mn.Nmu mmHuHHHns «0
oumaHumouumom

Ammo.m HNnm.Nm ow.wmwn om.mnH¢ om.mNnm m¢.m¢+ n~.w¢- muHsmou uuou no
oumawummnmum

mmmm.m NNmn.¢m wn.¢new mm.m¢o¢ om.wNmm no.mm+ mN.om- moHuHHHns «0
oumfiwumonoum

GOHumH>mo mocmHum> Hmuou nHuHm whoa nHun‘ axon
oumonmum

ham"...

‘|,l
l1|1

mm9<zHHmm m>HHommmmm mm? mo mz<mz

HHH> mqm<9

.w

 

 

nz<mu 20mm «945mmm Emma ho mm9<£HHmm onH<HmmmMMHzHHmom nz< onH<HmmmMmHszmm
mo 92¢ NHHHHm< ho mMH<ZHHmm onH<HmmMMMHZHHmom nz< 20HH<HmmmmmHszmm mo NHHAHm<HM<>

52

 

 

 

Noo. menn.H muHsmou umou mo mouuaHumouumon :uHa mmHuHHHnm mo ouwaHumouumom
uuu mNmo.H aanmou umou no moumEHumououn guHa moHuHHHns mo oumaHumououm
Noo. ame.m muHsmmu ans» «0 moumaHumouumon :uHa muHsmmu umou mo oumaHumououm
Noo. wwNH.N amHuHHHnm mo oumEHumouuuon :uHs umHuHHHnm we oumEHumououm
IllwmmmMMwMQMHm m memHummEoo woumnfiaocma u Houwuoesz

 

 

mHHDmMM Emma mo mmH<zHHmm onH<HmmmmmHZH

u Hmom Gz< ZOHH<memmmHZHmmm nz< mMHHHHHm< mo mMH<ZHHmm onH<HmmmMMHZHHmOm 92¢ mMH<ZHHmm
onH<HMMMMMHszmm mbomu A<HOB mo monH<ZHmzoo «DOHm<> mom moz<HM<> mo muz<loH20Hm mo mammaum

xH mHm<H

 

 

Noo. HNce.H aanmou nos» 90 oumEHumouumon nuHa moHuHHHnm mo ouuawumouumom

 

 

unu wNNo.H muHsmou umuu mo oumBHummuoun Hugs moHuHHHnu mo oumSHumououm
Noo. momm.m mUHsmou upon 90 sumaHumonumom saws muHsmou upon «0 oquHumoumum
Noo. quH.N moHuHHHnu mo sumaHumo-umon nuHa moHuHHHnm mo oumEHumououm
mocmoHMHnwww h :OHumcHnEoo uommmmmmmwauuoumuossz

 

 

mMH¢SHHmm ho m2¢mz 92¢MU 2029 mHHDmmm Emma 90 mMH¢EHHmm 20HH¢HNMNMMHZHHmom 92¢ ZOHH¢HmmmMMHZHMMm 92¢
mMHHH9H9¢ mo mMH¢ZHHmm ZOHH¢HmMmmmHZHHmom 92¢ 29HH¢99992982H929 .mwom mom NUZ¢HM¢> mo wo2¢UHthuHm mo mammanm

 

 

Hx mqm<a
3
5
onm.N wmmm.w manmou one» «o oumEHumouumom
Nmmm.m quN.¢H moHuHHHnm mo ousBHumsuumom
mmm¢.m oeow.mN muHsmou away no oumaHumonoum
mqmm.m NHmo.om mmHuHHHns mo mumaHumu-oum

 

cOHumH>m9 oumocmum ouanum>

 

 

mmH¢EHHm9 ho m2¢mz 92¢mu 929 2099 mHHDmmm Emmy mo m98¢2HHmm 20HH¢HmmmmmHZHHmom 92¢
ZOHH¢HmmmmMHZHmmm 92¢ MHHHHm¢ ho mmH¢ZHHm9 20HH¢HmmmmmHZHHmom 92¢ 20HH¢HmmmmmH2Hmmm .mwom m0 MHHAH9¢HM¢>

2 999¢H

54

the same page shows the results of testing the significance of the changes
in variance. After test interpretation, the boys had less variance in
both the estimates of abilities and in the estimates of test results,
differences significant at a highly significant level of confidence. No
significant difference was found between the variances of the preinterpretation
estimates of abilities and the preinterpretation estimates of test results.
However, after test interpretation, the boys had less variance in the
estimates of test results than in the estimates of abilities, a difference
significant at a high level of confidence.

Table XII on page 55 indicates the variability of the girls'
preinterpretation and postinterpretation estimates of abilities and of
test results as computed from the means of the total group, while Table XIII
on the same page shows the results of testing the significance of the
changes in variance. After test interpretation, the boys had less variance
in both the estimates of abilities and in the estimates of test results,
differences significant at a highly significant level of confidence. No
significant difference was found between the variances of the preinter-
pretation estimates of abilities and the preinterpretation estimates of
test results. However, after test interpretation, the girls had significantly
less variance in the estimates of test results than in the estimates of
abilities.

Table XIV on page 56 indicates the results of comparing variances
from grand means of various combinations of boys' and girls' preinter-
pretation and postinterpretation estimates of abilities and of test results.

Although the boys had a tendency toward less variance in both the pre-

55

 

 

NO. d3m...~ mafiflmmh U90“ NO flugflumfllumOQ 5U“? QU%U._UHHD¢ HO OugdUGOIUQON

 

 

 

--- NmHH.H manmou use» «0 ouuSHumououm saws moHuHHHnm mo oumaHumonoum

Noo. numo.m muHsmou umou mo ouuaHumouumon suHs uanmou umuu mo oumEHumououm

Noo. wNON.N moHuHHHnm mo oquHumouumon saws moHuHHHHw «0 oumEHumoumum

monmonchHm h noHumcHnaoo woumcHaonmanuoumumssz
mma<299mm mo

m2<mz 92¢MU 2022 mHHDmmm HmmH mo mmH<ZHHmm 20HH¢H9292992HHm09 92¢ 20HH¢892929H2H929 92¢ mMHH99H9¢
mo mma¢2HHmm 20HH¢HmmmMMHzHHmom 92¢ 2099¢HN¢922H2H929 .mHMHU 169 902¢HM¢> ho 992<UH9HZUHm mo mammaam

 

 

9992 99949
nomm.m moo¢.HH auHsmou umsu mo mumaHumouuuom
mm0~.¢ o¢H~.uH moHuHHHns mo ouuaHuaouumom
ONNm.n smoo.mm manmou umou mo ouuaaumououm
mm¢~.o oHso.mm mmHuHHHnm Ho ouuaHuou-mum

 

noHumH>o9 pudendum oocmHum>

 

 

mmH¢ZHHm9 mo m2¢mz 92¢du 92H 2629 mHHDmmm Emma 90 mmH¢2HHmm 20HH¢HmmmmmHZHHmom 92¢
ZOHH¢HMMAMNH2H929 92¢ NHH9H9¢ ho mmH¢2HHmm 20HH¢89292992HHm09 92¢ 20HH¢HN¢92982H929 .mAMHU ho NHHAH9¢HM¢>

9H2 999¢H

56

 

 

oH. mmmN.H muHSmou away no oumfiHummuuwom .mmon nqu muHsmou umou mo ousaHumonumon .mHuHo
u- oNeN.H mmHuHHHnm mo sumaHumouumoa .mxon nuHB aoHuHHHnu mo oumaHumouumon .mHuHU
.. oomH.H muHSmuu ummu we oumEHumououn .mzon nuHa muHsmou umou mo oumEHuuuuoun .mHuHu

2. $3; 3333.. no 3333-....3 .98.. 53 8333.. no 3333...: .338

 

mocmonchHm m mGOmHumnEou acumcHEoco9uu0umumezz

 

 

mmH¢ZHHmm ho m2¢mz 92¢MU 2029 mHHDmmm HmMH ho mmH¢ZHHm9
onH¢HmmMMMHzHHmom 92¢ ZOHH¢992929H2H929 92¢ meHHHH9¢ mo mmH¢ZHHmm 20HH¢HM¢92982HHm09 92¢
20HH¢89292982H929 .mHmHU 92¢ .mwom ho m2099<2992¢0 mDon¢> mom 902¢Hm¢5 ho 902¢UH9H20Hm ho mHmMHuh

>92 999¢H

57
interpretation estimate of abilities and the postinterpretation estimate
of test results, the differences did not reach an acceptable level of

confidence. No significant differences between the sexes were found.

IV. VARIABILITY OF ESTIMATES FROM SEX.MEANS

Table XV on page 58 indicates the variability of the boys' pre-
interpretation and postinterpretation estimates of abilities and of test
results as computed from the means of the sex group, while Table XVI
on the same page shows the results of testing the significance of the changes
in variance. After test interpretation, the boys had less variance in
both the estimates of abilities and in the estimates of test results,
differences significant at a highly significant level of confidence. No
significant difference was found between the variances of the preinter-
pretation estimates of abilities and the preinterpretation estimates of
test results. However, after test interpretation, the boys had significantly
less variance in the estimates of test results than in the estimates of
abilities.

Table XVII on page 59 indicates the variability of the girls'
preinterpretation and postinterpretation estimates of abilities and of
test results as computed from.the means of the sex group, while Table
XVIII on.the same page shows the results of testing the significance of
the changes in variance. After test interpretation, the girls had less
variance in both the estimates of abilities and in the estimates of test
results, differences significant at a highly significant level of con-

fidence. No significant difference was found between the variances of

58

 

 

No. «Nem.H muHsnou away no ouuaaumouumoa sud: uoHuHHHns mo ousfiauaonuuom

 

.- «HHo.H uuHsmou uses no sumaHumonoun nag: moHuHHHou mo suuaaumououm

Noo. mmoN.m muHsmsu umou mo sumaHumonumom and: oanmou umou mo sumsaumousum

3o. Sid 3323.. no 3233-38 :3: 3333.. no 3333-03
monsoHMHanm m . noHumaHoaoo uoumcaaononnuoumuoanz

 

 

mma¢SHHmu .mwom mo m2<mz.zomm mHHDmmm Emma 90 mMHEZHHmm 20HH¢992929H2HHm09 92¢.20HH¢99292992H929 92¢
mMHHHHHm¢ mo mma<2HHmm 20H9¢Hmmmmm92HHmom 92¢ 20HB¢99292992H929 .mwom mom 902¢Hm<5 mo 902¢UH9H2UHm ho mammaum

 

 

 

H>N 99949
NONo.m oHNH.m 0N.o¢HH mN.u nanmou umou no ouanumouumom
womn.m mwoo.¢H oo.mmNH nN.u moHuHHHnu mo ouuaHumouuuom
amne.n oops.m~ mw.o~am oo.o muHamau uuuu mo mumsHumm-mum
wa¢.n NONH.om so.mcmm on. u moHuHHHns mo sumaHumoaoum
noHumH>on 22235 023.32, Now H. W

 

 

mMH<ZHHm9 .mwom mo m2¢mz 928 2629 mHHDmmm Emma 90 mmadxHHmm 20HH¢99292992H8m99 92¢
20HH¢992929H29929 92¢ MHHAH9¢ mo mma<2HHmm ZOHH¢HmmmmmHZHHm09 92¢ 20HH¢99292992H929 .muom ho NHH9H9¢HM¢>

>2 999¢H

59

 

 

Noo. mmeo.H manmou umou mo oumaHumouumon so“: moHuHHHnm mo oumBHumouumom

 

nu oqu.H muHsmou umou mo oumawumonoun nuwa aoHuHHHom mo oumaHumououm

Noo. Hmmm.N manmou umou mo oumEHumonumoa suwa muHsmou umou mo oumEHumououm

8o. 23S 3:33.. mo magnumobmoa 53 .3333. no 3333.....3
monsonchHm m GOHumcHnEoo Housewaoconuuoumuoasz

 

 

mma¢zHHmm .mHMHU mo m2¢mz_2029 mHHDmNM Emma 90 mmH¢ZHHmm ZOHH¢HmmmmmH2HHm09 92¢ 20HH¢992929H2H929 92¢
mmHHHHHm¢ mo mma¢299mm 20HH¢HmmmmmHZHHmom 92¢ ZOHH¢HMMAM992H929 .mAMHU mom 902¢Hm<> ho 992¢UH2H20Hm mo mHmMHnm

 

 

 

I "11

HHH>x 9Hm<a
Noam.n oqu.HH om.mmmH Ho. - «UHsmsu umou mo mumsHumm-umom
no¢m.¢ ooew.mH cm.Hq- HH. + moHuHHHnm mo mumsaumm-uwom
mmm~.m mmHn.mm mm.mmmm 0H. +. muHamou away no ouqaauma-oum
mNoN.o oo~e.mm so.m~ms oo.a- moHuHHHna mo sumsHumo-aum
O w>m Md 0 0096 w
c Hu H n v as am an > Nva am“

 

 

mMH¢ZHHm9 .mHMHU mo m2¢mz 92H 2022 mHHDmmm Emma mo mmH¢ZHHmm 20HH¢99292992HHm09 92¢
onH¢HmmmmeZHmmm 92¢ VHHHHm¢ mo mMH¢ZHHmm 20HH¢HM¢9¢MH2HHmom 92¢ 20HH¢H2292982H929 .mHMHU ho NEH999¢HM¢>

HH>N 999¢H

60
the preinterpretation estimates of abilities and the preinterpretation
estimates of test results. However, after test interpretation, the girls
had less variance in the estimates of test results than in the estimates
of abilities, a difference significant at a highly significant level of
confidence.

Table XIX on page 61 indicates the results of comparing variances
from the appropriate sex means of various combinations of boys' and girls'
preinterpretation and postinterpretation estimates of abilities and of
test results. Although the boys tended toward less variance in all four

estimates, the differences did not reach an acceptable level of confidence.
V. MEANS OF THE HIGH AND LOW ACADEMIC APTITUDE GROUPS

Subjects were divided into high, middle, and low academic aptitude
groups on the basis of their total A.P.T. score in order to investigate
the relationship of academic aptitude to accuracy of estimation. Those
whose scores fell at or above .6 standard deviation above the mean of
the study group were put into the high academic aptitude group, and those
whose score fell at or below .6 standard deviation below this mean were
placed in the low academic aptitude group. Table XX on page 62 indicates
the results of comparing the means of these groups for significance of
difference. After test interpretation, both the mean estimate of abilities
and the mean estimate of test results of the high academic aptitude group
were significantly higher. After test interpretation, both the mean estimate

of abilities and the mean estimate of test results of the low academic

aptitude group were significantly lower.'

61

 

 

 

0H. mHmN.H manmou umou mo ouuaHummuumon .nhoo
Sufi: manmou umou mo oumEHumoaumom .mHuHu

oH. Hamm.H moHuHHHom mo oumaHumonumon .m%oa
no“? mmHuHHHow mo ouuaHumonumoa .mHuHo

oH. mmNH.H manmou use» no oumEHumoaoun .mhoo
zuwa manmou umou mo oumEHumououq .mHuHu

oH. ¢NNN.H moHuHHHom mo oumEHummnoum .mhoo

cows msHuHHHnm mo oumEHumonoun .mHuHu

oocmonchwm m wnOmHumnaou nouchEoconuuoumuoanz

 

 

mmaafiemm .335 an.
no mg may 92 $3.55.». .38 so mzfia m5. some 3ng Say so mmagsmu

20HH¢99292992HHm09 92¢ 20HH¢89292292H929 92¢ MMHHHHH9¢ ho mMH<2HHmm 20HH¢89292992HHm09
92¢ ZOHH¢99292982HHM9 .mHmHU 92¢ .mwom mo mZOHH¢2Hmon mDoHM¢> mom 902¢Hm¢> mo mo2¢UHMH20Hm mo mHmmHuh

2H2 999¢H

62

.nsouw hosum Hmuou any we asuoom .H.m.¢ Huuou mo nova osu Scum :OHumH>oo venomous o.

aoHoo no um HHom «spoon .H.m.¢ HmuoH muons muuofionm owoau mo vomomaou aw nsouw monuwumm oaaoomom 30H
may .nnouw honum Hmuou on» no mouoom .H.m.¢ Hmuou ecu we come o£u_aoum cOHuuH>oo ousonmum o. o>oom
no as HHom mououm .H.m.¢ HmuOH smog; muuonnsm among mo oomonaoo mH nnouw sonuwuds oflaoomos :an 059%

 

 

 

Ho. ommufi + mmmm. 3N + Hmém 3.2” madam.“ ummu mo mansions
nacho uwfioouo¢.3oa
mo. ommm.N + qu.H NN.N + om.om NN.mm moHuHHHnm mo mmumaHumo
nzouu ugfimvmo¢ BOA
Ho. nmon.m - oenw. NN.m n om.Nm «H.aN manmou umou mo moumawumo
macaw uwaooso¢ anm
Ho. mmmH.m u ommm. om.N . -.Nm Hm.mN moHuHHHom mo moumaaumo
ozone owfimomo¢ swam
xuw‘ amen. use:
ouanHchmHm u 1.18 9 nowumuoun :oHumumua
uuounwumom unsunHoum

 

 

¥9Domu m99HHHM¢ UHZM9¢U¢ 309 92H 92¢ 92020
H9DHHHM¢ 0H299<U¢ 20H2 92H 90 mHHDmmm Emma 90 mmH¢EHHm9 onH¢HmmmmMBZHHmom 92¢ 2099¢Hfl¢929fi29929
ho m2¢mz mo 92¢ mmHHHHHm¢ mo mmH<2HHmm 2098¢892922H29Hm09 92¢ 20HH¢89292292H929 ho m2¢mz mo 20m92¢mzoo

22 299¢H

63
VI. MEANS OF THE OVER-ESTIMATORS AND UNDER-ESTIMATORS OF ABILITY

Subjects were also divided into three groups on the basis of
their degree of accuracy in initial estimates of their abilities. Those
whose preinterpretation estimates of ability fell at or above .6 standard
deviation above the mean of estimates of ability of the total study group
formed the over-estimators of ability group. In a similar fashion, those
whose preinterpretation estimates of ability fell at or below .6 standard
deviation below the mean of the preinterpretation estimates of abilities
composed the under-estimators of ability group. Table XXI on page 64
reports the results of testing the means of estimates of these groups
for significance of change. After test interpretation, both the
mean estimate of abilities and the mean estimate of test results of
the over-estimators of ability group were significantly lower. After
test interpretation, both the mean estimate of abilities and the mean
estimate of test results of the under-estimators of ability group were

significantly higher.
VII. MEANS OF THE OVER-ESTIMATORS AND UNDER-ESTIMATORS OF TEST RESULTS

Finally, over-estimators and under-estimators of test results were
identified on the basis of their initial estimates of test results.
Those whose preinterpretation estimates of test results fell at or
above .6 standard deviation above the mean of preinterpretation estimates
of test results of the total group were placed in the over-estimators of

test results group. Those whose preinterpretation estimates of test

64

.nnonw spawn Huuou can

we huHHHom mo moumEHumouonm 0:» mo amps man Eonw conumw>ov vnmonmum c. soHon no no HHow huHHHnm mo
moumfiHumouonm omo£3 muoohnnm omoeu mo vmmonaoo mH anonw huHHHnm mo unoumEHumounoon: 02H .nsonw monum
Hmuou can 90 aunHHnm mo moumEHumouonn can «0 some can Bonn nOHuuH>oo vnmonmua o. o>oom no no HHow
muHHHom mo mmumEHumouonn muons unconcnm omosu mo ommooaoo an anonw huHHHom mo anoumEHumunno>o ones

 

 

Ho. oeom.e - Homm. mm.¢ . om.Nm mm.uN muHsmmn ammo mo
moumEHumo .muHHHom
mo mnoumanumounooab

Ho. H~n~.q - «soo.H m~.¢ - Ha.Hm Ns.n~ aunnnna no
moumEHumo .zuHHHom
wo mnoumEHumo:nooaD

no. ¢¢m0.N + Mona. om.H_+ mn.mm m¢.~m muHsmmn umou mo
moumBHumo .muHHHom
mo mnousanumoano>o

no. s-q.m + neso.n no.m + on.nm Hs.a¢ aunnnsm no
moumEHumo .quHHom
mo muougwumulums

 

2-2 cmmz smog
cosmoHMchHm u I.Id 9 nOHumumnm :oHumumnm
unsunaumom -nounHonm

 

 

*NHHHHm¢ mo mmOH¢2HHm9nmm92= 929 no 92¢ MBHHHm¢
mo mMOH¢ZHHmmu29>O 928 90 mHHDmmm Emma 90 m9H¢ZHHmm 20HH¢HmMMMMHzHHmom 92¢ ZOHH¢822922929229
mo m2¢mz 92¢ mMHHHHHm¢ mo mMH<ZHHmm 20HH¢99292992HHm09 92¢ 20HB¢992922H2H229 mo m2¢mz ho mZOmHm¢mZOU

H22 999¢H

65
results fell at or below .6 standard deviation below the mean of
estimates of test results formed in the under-estimators of test results
group. Table XXII on page 66 contains the results of testing the means
of estimates of these groups for significance of change. No significant
difference was found between the mean of the preinterpretation estimate
of abilities and the mean of the postinterpretation estimate of abilities
of the over-estimators of ability group. However, after test inter-
pretation the mean estimate of test results of the over—estimators of
test results group was significantly lower. For the under-estimators
of ability group, boththe postinterpretation estimates of abilities
and the postinterpretation estimates of test results were significantly
higher.
VIII. MEANS OF THE SUB-GROUPS COMPARED TO GRAND MEANS OF THE TOTAL
STUDY GROUP

The next step in the analysis of the data was to compare the
means of the above-mentioned sub-groups with the means of the total
study group. Table XXIII on page 67 shows the comparison of the means
of the high academic aptitude group with the means of the total group.
All four mean estimates of the high academic aptitude group were
significantly lower than were the mean estimates of the total group
in the study. For the high academic aptitude group, the mean postinter-
pretation estimates of abilities and of test results were very close
to the points of maximum accuracy.

Table XXIV on page 67 shows the comparison of the means of the

low academic aptitude group with the means of the total study group.

66

.msonw moans Hauou on» 90 some ecu sonw nouuoH>oo oneonnum o. 3oHon no no HHom snoop manmon
umou mo moumanumanona omega muoonosm omonu mo oomonaou mH 990nm manmsn upon «0 mnousanumounoonn 09H .nnonw
moans Hsuou can no uanmon upon 90 moumaHumouonm can «0 some can Bonn noHumH>oo onmonmum o. o>oom no um HHom
muHawsn upon 90 mmumEHumoumnm smog: muoononm among mo omwomsoo mH 930nm manmon noon «0 mnoumaHumounm>o seas

 

 

 

Ho. mNNN.m u MMNm. om.< J Ho.Nm HN.NN manmon umou
mo moumBHumo .muHsmon
away 90 mnoumanumounoon9
Ho. mmom.N . comm. Hm.N . co.Hm mH.mN uoHuHHHos
mo mounaHumo .muHsmon
upon 90 mnoumaHumounoocs
Ho. hmoa.¢ +. mHHm. mm.¢ + mm.mm wN.o¢ muHsmon away
no moumEHumo .muHamon
ummu mo anoumBHumonno>o
-- omom.H + mNom. ow.H + NH.wm oo.o¢ muHHHom «0
mmumEHumo .manmon umou
mo unoumaHumo-nm>o
2-2 smoz new:
monmonchHm u I.Id 9 nOHuwuonm nonumuonm
unsunnumom unsunnonm

 

 

¥m899m22 Emma 90 mMOH<ZHHmmumm9ZD 92H
90 92¢ mHHDmmm HmmH mo wMOH¢ZHHm9uMm>O 92H 90 mHHDmmm Emma 90 mme¢ZHHmm 20HH¢992929H29Hm09 92¢
20HH¢992929829929 92¢ mMHHHAHm¢ mo mmH¢ZHHmm 2098¢992929929Hm09 92¢ 2099¢892922H2H929 ho 92¢22 ho mzomHm¢mzoo

HH22 999¢H

67

 

 

 

mo. omwm.N u 890m. «m.H u Hm.¢m 8n.nm manmon upon
90 oumaHumouumom

Ho. NNmm.N u meo. no.9 . on.om hm.¢m moHuHHHnm mo
mumaaumonunom

Hoo. Nme.¢ u mHHw. H¢.m u ow.8m mm.mm manmon you»
90 oumaHumOnonm

Hoo. «ONH.m u mwa. hm.¢ a NN.mm mo.¢m moNuHHHnm mo
mumaHumoumnm

009.339:me u MHMm 9 anonolwonuwu n¢ 935...... on... 99

awesomo< 309 ozone Hmu08

 

 

macaw 8998m 9¢808 228 20 m2¢22 228 2892 macaw 09229¢U¢.309 228 20 mz<22_228 mo 20m92¢9200

 

 

 

>H22 292¢8

Hoo. omno.m + mmNm. HN.H +. em.Nm nm.mm muHsmon umou
mo sumawumouumom

900. N09.— ..v + 9092. 09.N + hh.Nm swium 90999999d mo
oumEHumouumom

Hoo. comm.m + mmmw. Hm.¢ + 8H.mN mm.mm muHSmmn umou
MO Ouaflumwlmhm

900. «3%. + 0992. cm I» + 90.0N morqm 909999.992 mo
sumaHuaouonm

OUGQUHwflwam u Mama . Q 990.90 OUSUHUQ< 229.5% USU 9.“

afiemomo¢ an2 930nm Hmu08

 

 

99020 89982 9¢808 228 29 m2¢22 228 2893 99020 09229<U¢ 2092 228 mo 22¢22 228 29 20m92¢mzcu

99922 292¢8

68
All four mean estimates of the low academic aptitude group were significantly
higher and farther from the point of maximum accuracy than were the mean
estimates of the total group in the study.

Table XXV on page 69 shows the comparison of the means of the over-
estimators of ability group with the means of the total study group.

All four mean estimates of the over-estimators of ability group were
significantly higher and farther from the point of maximum accuracy
than were the mean estimates of me total group in the study.

Table XXVI on page 69 shows the comparison of the means of the
under-estimators of ability group with the means of the study group. All
four mean estimates of the under-estimators of ability group were significantly
lower than were the mean estimates of the total group in the study. The
mean postinterpretation estimates of abilities and of test results of the
under-estimators of ability group were very close to the points of maximum
accuracy.

Table XXVII on page 70 shows the comparison of the means of the over-
estimators of test results with the means of the total study group. All
four mean estimates of the over-estimators of test results group were
significantly higher and farther from the point of maximum accuracy than
were the mean estimates of the total group in the study.

Finally, Table XXVIII on page 70 shows the comparison of the means
of the under-estimators of test results group with the means of the total
study group. All four mean estimates of the under-estimators of test
results group were significantly lower than were the mean estimates of

the total study group. For the under-estimators of test results group,

69

 

 

 

Hoo. 0N00.m + ween. 8N.H + om.Nm 9n.mm mUHnmon umou
mo oumaHumouumom

Hoo. Nmnm.n + meow. 0H.m + HN.Hm mm.¢m moHuHHHnm mo
oumEHumouumom

Hoo. meow.o + maHm. mm.m +. ha.NN nm.mm manmon use»
no oumaaumouonm

Hoo. deo.n + Hosm. MN.N +. No.9N mo.¢m moHuHHHnm mo
oumEHumoumnm

ooamonchHm u 2-2 9 moonwjzuHHwnm mo monum may 99
I.Id

mnoumaaumsunonea

anonu Hmu08

 

 

99022 89982 A¢808 228 20 mz¢22 228 2893 99029

889292¢ 20 mMO8¢ZH8m2u22929 228 29 22<22 228 20 20m92¢2290

 

 

 

9>22 222¢8

Hoo. om¢.m . Hwom. om.9 - mm.nm 8m.mm manmon upon
90 sumaaumouumom

Hoo. Hum¢.¢ . nch. aw.N u 08.nm sw.¢m moHuHHHcm «0
sumawumouumom

Hoo. sumo.e . womm. am.m a me.sm n¢.mm muHsmon upon
90 oumEHumouon9

Hoo. HoHo.u . nmom. om.o u H¢.H¢ no.8m moauwHHnm mo
sumaHumouonm

mochHanme u MHMu 9 [moonwlmuHHHnm mo madam man c9

nnoumaHumounm>o anonu Hmu08

 

 

99020 89982 A¢808 228 20 m2¢22 228 2893 99020 889992¢ mo mMO8<298m2n22>o 228 20 mz¢22 228 20 20m92¢92oo

>22 292¢8

70

 

 

 

Hoo. mNHm.¢ hmcm. om.H +. Ho.Nm hn.mm manmmn ummu
mo oumaqumouumom

Hoo. mNo.o mNmm. HN.m + oo.Hm nm.¢m mmHuHHanm mo
oumEHumouumom

So. 3: .9 Now. 3.0 + HTS 3.3 3153 ummu
mo oumaaumousnm

Hoo. oHoo.o Hwa. on.m +. nH.mN no.8m mmHuHHHcm mo
oumawumouonm

monsonacme n MWMm 9 moono muHsmom umo8 .NUSum may :9

mo anousawumounooaa moonu Hmuo8

 

 

99020 89D8m 9¢808 228 20 92<22 228 2893 99020

m899m22 8928 20 22084298m2n22929 228 20 22¢22 228 20 20m92¢2200

999>2 292¢8

 

 

 

Hoe. Hne~.q - oHHm. mH.~ - ma.nm nm.mn munsmau smug
MO Ouawumolumom

Hoe. mamm.n - Anne. om.m - an.mm aw.em monunnnnu no
sumaaumoaumom

Hoo. «Ham.n - snow. mm.s - m~.oe ma.mm munsmon same
no oumEHumsnonm

Hoo. nm-.o - omsm. Hq.m - oo.o¢ mo.qm monunnnnu no
sumanuusuonm

uuauonwnamnm u mum. a macho mannnn< no mssum was an

mnOumEHuuounm>0 mson0 Hmu08

 

 

99020 89=8m 2¢808 228 20 m2¢22 228 2893 99020 m899m22 8m28 20 9208229822nm2>0 228 20 m2¢22 228 20 20292¢9200

99>2 292¢8

71

the mean postinterpretation estimates of abilities and of test results
were very close to the points of maximum accuracy.

IX. PROPORTIONS OF BOYS AND GIRLS IN THE TOTAL STUDY GROUP AND IN THE

VARIOUS SUB-GROUPS
The total study group and each of the six sub-groups were examined
to determine the significance, if any, of the pr0portions of boys and
girls. This data appears in Table XXIX on page 72. No group in the study,
including the total group, was found to differ significantly in the
proportion of boys and girls, from what would ordinarily be expected.
X. VARIABILITY 0F ESTIMATES OF THE HIGH AND LOW ACADEMIC APTITUDE GROUPS
FROM THE GRAND MEANS OF ESTIMATES
As noted previously, the major statistical technique in the study

was to be the analysis of variance of the preinterpretation and post-
interpretation estimates of abilities and of test results. Table XXX
on page 73 presents the variances and standard deviations of the four
estimates of the high academic aptitude group computed from the total
group's means. Table XXXI on the same page contains the results of testing
these variances for significance of change. After test interpretation,
there was a highly significant decrease in the variances of both the
estimates of abilities and the estimates of test results for the high
academic aptitude group. No significant difference in variance was found
between the preinterpretation estimate of test results and the preinter-
pretation estimate of abilities. After interpretation of test results

there was less variance in the estimates of test results than in the

72

TABLE XXIX

CHI SQUARE TESTS OF SIGNIFICANCE OF THE PROPORTIONS OF BOYS AND GIRLS IN

THE VARIOUS GROUPS OF THE STUDY

 

 

Total Group in the Study

 

 

 

 

 

 

 

 

 

 

 

 

 

 

observed expected X2 Significance
boys 125 122
girls 119 122 .
244 244 .1474 --
High Academic Aptitude observed expected 12 Significance
Group "“’
boys 31 34
girls '_35 32
66 66 .5459 ~-
Low Academic Aptitude observed expected X2 Significance
Group
boys 38 40
girls 40 38
78 78 .2052 --
Over-Estimators of observed expected X2 Significance
Ability
boys 35 38
girls 39 36
74 74 .4868 --
Under-Estimators of observed expected X2 Significance
Ability
boys 39 36
girls 32 35
79 79 .5017 --
Over-Estimators 0f observed __expected X2 Significance
Test Results
boys 37 40
girls _5; 39
79 79 .4557 --
Under-Estimators of observed expected X2 Significance
Test Results
boys 40 41
girls 40 39
80 80 .0499 ""

 

 

 

 

 

 

Noo. SEN manmon umou mo ougaumouumom 2a.“: «0.3.99.2? mo oumaHumouumom

 

E. . 22.0.9 moHuHHaou mo ouaaumouonm EH93 manmon umou mo mumaaumoaonm

moo. mama.m muHamon smug we oumeHumo-umoa nuns uuHsmon name no mumsnumu-onm

Noo. m¢mm.m moHuwHanm mo oumEHumonumon an? «0.3.3.23 mo uumEHumoumnm
monmonEfim 2 chmHnmnfioo noumcHEoco9unoumno852

 

 

228329822 .20 22422 92¢20 202.9 28.99222 8228
.20 228329822 2098¢8229228298209 92¢ 2098¢822922929229 92¢ 22989292¢ .20 228329822 2098¢8229228298209 92¢
2098¢822922829229 2.99020 299898m¢ 09229¢o¢ 2092 20 22098¢292200 29092¢> 202 202492¢> .90 20259292092 .20 28228-2

 

 

 

Hxxx mam¢n

3

7
osmo.N onmm.e ¢~.oaN No.o¢ - uuHsmun use» no onusnumouumom
m~o¢.m amaH.~H mH.mow ~¢.¢mH - monuaHHnu no onusHumo-uaom
236 33.3 8.83 2.3m .. 3963 new» no 3233-39
$2.0 $3.3 «9:33 8.2m .. 3333.. no 3333-39

cOHumH>o9 onmpcuum moannm> Now How

 

 

99020 2998989¢ 09229¢U¢ 2092 228 202
22839822 2>98029222 228 .20 22.222 92¢20 2022 28.99222 8228 .90 228229822 2098¢8229228298209 92¢
2098¢822922829229 20 92¢ 889992¢ 20 228329822 2098¢8229228298209 92¢ 2098¢822922829229 .90 8899925239

222 292¢8

74
estimates of abilities, a difference significant at a highly significant
level of confidence.
Table XXXII of page 75 presents the variances and standard
deviations of the four estimates of the low academic aptitude group,
while Table XXXIII on the same page reports the results of testing these
variances for significance of difference. After test interpretation,
the low academic aptitude group had a highly significant decrease in
variances of estimates of both abilities and of test results. No
significant differences in variance were found between either the pre-
interpretation estimates of abilities and the preinterpretation estimates
of test results, or between the postinterpretation estimates of abilities
and the postinterpretation estimates of test results.
XI. VARIABILITY 0F ESTIMATES OF THE OVER-ESTIMATORS 0F ABILITY AND OF
THE UNDER-ESTIMATORS OF ABILITY FROM THE GRAND MEANS OF ESTIMATES
Table XXXIV of page 76 presents the variances and standard deviations

of the four estimates of the over-estimators of ability group, while
Table XXXV on the same page reports the results of testing these variances
for significance of difference. After test interpretation there was a
highly significant decrease in the variances of both the estimates of
abilities and the estimates of test results. No significant differences
in variances were found between either the preinterpretation estimate
of abilities and the preinterpretation estimate of test results, or between
the postinterpretation estimate of abilities and the postinterpretation

estimate of test results.

75

 

 

nu onmo.9 mu9amou umou mo ouuawumouumom :u93_mo2u29wnm mo ouuaaumouumom

 

u- m¢¢~.2 mu25mmu umou mo oumaaumo-ou2 :u23.m02u29«nu mo ouuaaumououm

Noo. «m2m.2 muHamou umou mo oumaflumouumon nu23 mu2=mou umou mo oumafiumououm

«co. mmmo.~ uuflufififinu mo uuuafiumo-umoa as“: manuafiflnw mo muuafiumo-uum
mocuofiwacmwm 2 mooflumaanaoo nouocwaoaonuuouwuoadz

 

 

228¢29822 20 22¢22 22¢20 2022 2822222 8228 20 228¢29822 2098¢8222228298202
22¢ 2098¢822222829222 22¢ 22989292¢ 20 228¢29822 2098¢8222228298202 22¢ 2098¢822222829222
2.22020 2228982¢ 09222¢0¢ 302 20 22098¢292200 22092¢> 202 202¢92¢> 20 202¢09292092 20 28228u2

 

 

 

999222 222¢8
mumo.¢ qmom.2~ 22.2229 qw.o02 + mu2=mou umou mo oumauumouumom
2290.2 o¢¢n.m~ 20.200N «9.222 + moauq9anm mo oumafiumonumom
N¢o¢.2 2q20.2¢ 09.mm22 om.¢om + mu93mou umou «0 oumaflumo-ou2
22¢9.~ 9¢mo.22 22.02mn 02.222 + mo2u292nm «0 mumswummuoum
mmmwmw>o2 vumvcmum oucwwum> Nvmw vmw

 

 

22020 2228982¢ 09222¢0¢ 302 228 202
228¢29822 2>98022222 228 20 22¢22 22¢20 228 2022 2822222 8228 20 228¢29822 2098¢8222228298202 22¢
2098¢822222829222 20 22¢ 289292¢ 20 228¢29822 2098¢8222228298202 22¢ 2098¢822222829222 20 289292¢92¢>

99222 222¢8

76

 

 

In 0022.2 mu29muu umuu mo oum82umoaum02 :u23 uo2u222nd mo uumfi2umuuun02

 

.i $2.2 325mm“ umou mo ouua2umonou2 523 32.22222? 20 038230-022

~00. -2~.N 32260.2 you» no 038230-302 5.23 muggy umou mo 39:33-32

moo. nom~.~ mo2u222nu mo ouuafluau-u.oa nuaa .o2u2222a mo ouuaqumu-uum
005322222222 2 303322300 “2332805933552

 

 

228¢29822 20 22¢22 22¢2U 2022 2822222 8228 20 228¢29822 2098¢8222228298202
22¢ 2098¢822222829222 22¢ 22989292¢ 20 228¢29822 2098¢8222228298202 22¢ 2098¢822222829222
2.22022 289292¢ 20 2208¢29822u22>0 20 22098¢292200 22092¢> 202 202¢92¢> 20 202¢UH292292 20 28228u2

 

 

 

>222 222¢8
N22m.¢ 2qow.o~ om.amm2 «2.222.+ mu2smou ummu «0 auuaHumu-umom
nmm0.2 0200.2N ~2.¢~m2 «2.22N +. «0222222m mo ouma2umonum02
2022.2 320.3 02.~2¢m 02.222 + 32.60.“ ummu mo 3333-022
02~2§ 2222.22 25.2022 8.8.2 + 9.8322223 20 ouua2umuumu2
602um2>m2 28.825922 «2.82.23, 22 W 2 W

 

 

22022 289292¢ 20 2208¢29822u22>0 228 202
228¢29822 2>98022222 228 20 22¢22 22¢2U 228 2022 2822222 8228 20 228¢29822 2098¢8222228298202 22¢
2098¢822222829222 20 22¢ 289292¢ 20 228¢29822 2098¢8222228298202 22¢ 2098¢822222829222 20 289292¢92¢>

>92NN 222¢8

77

Table XXXVI of page 78 presents the variances and standard
deviations of the four estimates of the under-estimators of ability group,
while Table XXXVII on the same page reports the results of testing
these variances for significance of difference. After test interpretation,
the under-estimators of ability group had a highly significant decrease
in variances of both the estimate of abilities and of test results. No
significant difference in variance was found between the preinterpretation
estimate of abilities and the preinterpretation estimate of test results
for this group. However, after test interpretation, the under-estimators
of ability group had less variance in their estimates of test results
than in their estimates of abilities, a difference significant at a
highly significant level of confidence.

XII. VARIABILITY OF ESTIMATES OF THE OVER-ESTIMATORS OF TEST RESULTS AND
OF THE UNDER-ESTIMATORS OF TEST RESULTS FROM THE GRAND MEANS OF ESTIMATES
Table XXXVIII of page 79 presents the variances and standard

deviations of the four estimates of the over-estimators of test results
group, while Table XXXIX on the same page reports the results of testing
these variances for significance of difference. After test interpretation,
there was a highly significant decrease in the variances of both the
estimates of abilities and the estimates of test results. No significant
difference in variance was found between the preinterpretation estimate

of abilities and the preinterpretation estimate of test results. For

the over-estimators of test results group, there was a tendency in the
direction toward more variance in the postinterpretation estimate of
abilities than in the postinterpretation estimate of test results, but

this difference did not reach the acceptable level of confidence.

78

 

 

NOO- w¢chM mUHfimUH 0200 NO flugflumflluuom Sufi? MUMUHHflfld MO Ougflummtumom

 

- 2220.2 mu2amuu umou mo ouma2umo-ou2 :u23.mo2u2222c mo ouua2umo-ou2
~00. «222.2 mu2smou umou mo oumfi2umo-um02 Saws mu2=mou umou mo ouma2umo-ou2
~00. 2222.2 mo2u2222¢ «0 ouma2umo-umoa nu23.mo2u2222w «0 ouma2umo-ou2

oocwo2w2cmwm 2 ch2uo=2naoo noumc2aocon-uououoasz

 

 

228¢29822 20 22¢22 22¢20 2022 2822222 8228 20 228¢29822 2098¢8222228298202
22¢ 2098¢822222829222 22¢ 22989292¢ 20 228¢29822 2098¢8222228298202 22¢ 2098¢822222829222
2.22020 289292¢ 20 2208¢29822-22222 20 220982292200 22092¢> 20 202¢92¢> 20 202¢09292092 20 28228-2

 

 

99>222 22248
2222.2 «222.2 «2.202 22.02 - mu2=mmu ummu mo ouma2umo-um02-
2222.2 2222.22 «2.2222 22.222 - mo2u222na «0 mums2umo-umom
0222.2 2022.02 0~.2222 22.222 - mu2smou umou mo ouma2umo-on2
2022.2 2222.22 «2.2222 22.222 - mo2u2222o mo ouua2umo-ou2
co2uw2>on wumwcwum mocm2um> Nwmw 2omw
.IHHHHNHP1 «nun

 

 

22020 289292¢ 20 2208¢29822-22222 228 202
228¢29822 2>98022222 228 20 22¢22 22¢20 228 2022 2822222 8228 20 228229822 2098¢8222228298202 22¢
2098¢822222829222 20 22¢ 289292¢ 20 228¢29822 2098¢8222228298202 22¢ 2098¢822222829222 20 289292¢92¢>

9>222 222¢8

 

79

02. 2222.2 mu2=mou umau mo ouma2umunumoa :u2RVuu2u222nu we uuua2umonumo2

 

u- 2000.2 mu292uu umwu mo uumE2umouou2 nu23 um2u2222¢ mo uumfi2umoumu2

200. 2222.2 mu2=mou ummu mo muua2umouumo2 :u23 uu2smou umou mo ouua2umouou2
woo. 222m.2 ao2u222am 2o ouma2umo-umoa nu2a mu2u2222u mo muma2umo-oum
moamo2M2Gw22 2 mc02uud2nsou wouQGHanmnuuouwH0552

 

 

228222822 20 22<2Z 22<2U 2022 2822222 8228 20 228222822 2028<8222
:228228202 22¢ 2028<822222822222 22¢ 22H8H2Hm< 20 228222822 208848222228228202 22¢ 20282822222828222
2.2202U 2822222 8228 20 2208<22822322>0 20 220884222200 220H2<> 202 202<H2<> 20 2024UH2HZUH2 20 2822812

 

 

 

22502 522.2.
2022.2 228.2 22.2222 22.22 + 3283 ummu mo 32.33-83
2222.2 2222.22 22.2222 22.022 + mm2u22222 «0 muma2umouumo2
2222.2 2002.22 22.2222 22.222 + mu2zmou umuu mo ouma2ummumu2
2222.2 2202.22 22.222m 22.2n2_+ au2u2222a 2o ouua2umo-oum
:02uw2>m2 vuwvcmum oocw2um> 22mw 2mw

 

 

22020 2822222 8228 20 220822H822u22>0 228 202
228228822 2>H8022222 228 20 22222 22222 228 2022 2822222 8228 20 228222822 2088<8222228228202 22¢
20284822222822222 20 22< W8H2Hm< 20 228228822 20H8<8222228288202 22¢ 2028<8222228ZH222 20 288282<H2<>

HHH>NNN 222<8

80

Finally, Table XXXX of page 81 presents the variances and standard
deviations of the four estimates of the under-estimators of test results
group, while Table XXXXI on the same page reports the results of testing
these variances for significance of difference. After test interpretation,
there was a highly significant decrease in the variances of both the estimates
of abilities and the estimates of test results. No significant difference
in variance was found between the preinterpretation estimate of test results
and the preinterpretation estimate of abilities. After test interpretation,
the under-estimators of test results group had less variance in their estimates
of test results than in their estimates of abilities, a difference significant
at a highly significant level of confidence.
XIII. VARIABILITY OF ESTIMATES OF HIGH ACADEMIC APTITUDE PUPILS COMPARED

TO THE VARIABILITY OF ESTIMATES OF LOW ACADEMIC APTITUDE PUPILS

Table XXXXII on page 82 shows the results of comparing the variances
of the four estimates of the high academic aptitude group with those from
the low academic aptitude group. No significant differences in variances
were found between either the preinterpretation estimates of abilities or
the preinterpretation estimates of test results of these two sub-groups.
However, after test interpretation the high academic aptitude group had
less variance in both the estimate of abilities and the estimate of test

results, differences significant at a highly significant level of confidence.

XIV. VARIABILITY 0F ESTIMATES OF OVER-ESTIMATORS 0F ABILITIES COMPARED TO
THE VARIABILITY 0F ESTIMATES OF UNDER-ESTIMATORS 0F ABILITIES

Table XXXXIII on page 84 shows the results of comparing the

81

 

 

moo. ooom.~ muaamou was» no oumawumouumoa sues nowadawnu mo mumsgumouumom

 

u- nH~H.H moguHAHnm mo oumawumououm nuaz muaamou umou mo ouuawuoououm
moo. cwqm.n muaamou umou mo oumawumouumoa so“: nuanmou umou mo ouuaaumonoum
moo. nmfim.~ mofiuqfianw mo oumawumouuoom now: mowuaafinw «0 oumaaumonoum

oucwofimwcwam m acowumaanaou uoumdaaoconupouwumadz

 

 

. mmH<ZHHmm mo mz<m2 nz<mu ZOMM «Haammm HmflH mo mMH<ZHHmM onH<Hmmm
uMMHzHHmom nz< ZOHB<Hm&NMMHZHmmm 92¢ mmHHHAHm< mo mMH¢ZHHmm onH<Hmmmmm92HHmom 92¢ ZOHH<HmmmmmHZHmmm
w.mnomu magnum“ Emma mo mmOH¢ZHHmmuMMQZD mo mZOHadszzoo WSOHM<> «Om muz<Hm<> mo moz<UHMquHm mo mammaum

 

 

 

HNNNN mqm<H
mom¢.~ m~m~.e ~n.wma oo.¢- . auasmuu umuu mo ouaeaumm-umom
a~m~.¢ wmm¢.~H ma.mom~ oo.om~ . moauaaano mo oumawumouumom
mm~m.o wwoa.m¢ o~.NHmm oo.am¢ - muasmou umuu «0 oumaaummuoum
mmmo.o aooo.¢¢ mo.o~mm oo.o¢¢ . moguwaano «0 muuaaumououm
COduwfixer thficwum OUGQHHw> Nmuw Cw

 

 

macaw mHADmmm Emma mo mMOH<ZHHmmnMMQZD mmH mom
mmH<zHHmm m>HHommmmm mmH mo mz<mz QZ<MU NEH Scam mHADmmm Emma mo mMH<2HHmu ZOHH¢HmmmmmHZHHmom 92¢
ZOHH<HmmmmmHszdm mo 92¢ MHHAHQ< mo mMH<zHHmN ZOHH<HMMAMMHZHHmom nz< ZOHH<HMMNMMHZHMMN mo MBHAHm<Hm<>

xxxx MAQ<H

82

 

 

woo. mwmw.¢ muaamou umou wo oumaaumouumom maouw opduaumw oaamvmoo swan
soak muaamou umou mo oumaaumouumon maouw ovauguao owewvmoo 30A

Noo. NoHH.~ moauaawno mo oumaaumouumoa spouauaw ofiaovwow
swan saga moauwaano mo oumaaumonumoa nsouw ovsuaumu ofiawvmow 304

u- nnoo.a muaamou umou mo oumaaumououa aaouw ovauwuna uafiovuoo 30H
:uH3 muaamou umou mo oumawumououm nsouw ovauaumm ousovmoo swam

nu n¢m~.H mowuaaanu mo oumaqumonoum maouw opsuauau anovuou
swan nuaz moauaaanu mo ouuaauwououm macaw ovauwumw oafiovoom 304

 

cosmofiwacmqm m mcoaumcwnaoo uoumcfiaoconuuouwuoaaz

 

 

mAHmDm MQDHHHmd UHZMQ<U< 30A mHHz mAHmDm mQDHHHm< UHZMQ<U<
mUHm UZHm<mzoo mHADmmm Emma m0 mMH<ZHHmm ZOHH<HmMmmMHZHHmom Dz< ZOHH<HMMMMMHszxm mom
mMHHHAHmd mo mMH<ZHHmm ZOHH<HmmmmezHfimom Q24 ZOHH<HNMNMMHZHMMN mow moz<Hm<> mo moz<onquHm mo mHmMHum

HHNNKN mqm<9

83
variances of the four estimates of the over-estimators of ability group
with those from the under-estimators of ability group. No significant
differences in variances were found between either the preinterpretation
estimates of abilities or the preinterpretation estimates of test
results of these groups. Although the under-estimators of abilities
group had a smaller variance in the post-estimate of abilities, the dif-
ference did not reach the acceptable level of confidence. After test
interpretation, the under-estimators of ability group had less variance
in the estimates of test results than did the over-estimators of ability
group, a difference significant at a highly significant level of con-
fidence.
XV. VARIABILITY OF ESTIMATES OF OVER-ESTIMATORS OF TEST RESULTS COMPARED
TO THE VARIABILITY OF UNDER-ESTIMATORS OF TEST RESULTS
Table XXXXIV on page 85 shows the results of comparing the variances
of the four estimates of the over-estimators of test results group with
those from the under-estimators of test results group. No significant
differences in variances were found between either the preinterpretation
estimates of abilities or the preinterpretation estimates of test results
of these two groups. Although the under-estimators of test results group
had less variance in the postinterpretation estimate of abilities than
did the over-estimators of test results group, the difference did not
reach the acceptable level of confidence. After test interpretation, the

under-estimators of test results group had less variance in their estimate

84

 

 

woo.

OH.

nwmo.m

mHm¢.H

oqmo.a

quo.a

muasmou umou mo mumswumOuumom msouw huwaanw mo muoumawumouuovaa
nuH3.muH=mou umou mo oumawumOuuooa macaw muaaanu mo ououoaaumouno>o

moquwaanw mo oumawumouumom macaw huwfianw mo muoumaaumo
uuovcs saga aoauaflanm mo mumsaumonumoa nsoum huwawna mo muouuafiumouuo>o

uuasmou umou mo oumfiaumououm anon» muwawnw mo muoumaaumouuo>o
saga muasmou umou mo oumaaumonoum msouw muaagnm mo muouuaqumouuovcz

moquaawnm mo oumawumououm asouw huaaanm mo ououmaauuo
unavas saw: wouuqaano mo ouuaaumououa anon» huwaanw mo muouuaaumoauo>o

 

cocoowmaawam

mGOHUgHQfi—OU HOUgHBOfiMA—IHOUNHGSZ

 

 

mMHHHAHm< mo mma<xHHmm onH<HmmmMMHZHHmom 92¢ onH¢HmmmmmHszmm mom moz<Hm<> mo moz<UHmHZUHm mo mammaum

NHHAHQ< mo mMOH<ZHHmmumHQZD mma MHHS NHHAHm< mo mMOH<ZHHmM
umm>o mmH Usz<mzoo mHADmmm Emma mo mm9<£HHmm onadfimMmmmHZHHmom 02¢ onfidHMMMMMHZHmmm mom nz<

HHHNNNN mum<s

85

 

 

moo. nnnn.m muadmou umou mo mumswummnumoa macaw muazmou umou mo muoumawumonuovcs
saws muaamou umou mo audawumouumon maouw muHSmou uwou mo muoumawumonuo>o

0H. mnaq.a moHuHHHnm_mooumafiumouumom macaw muasmou away no muoumaaummuumpcs
nova mofluwawnw mo oumawumouumom msouw muaammu umou mo muoumaaumonuo>o

s- moHo.H muasmou umou mo oumEHumo-oum macaw muasmou umou mo muoumafiumouuo>o sues
muasmuu umou mo oumawummtoua nsouw muHSmou umou mo mucquaumouuopc:

u- Nmoo.H mowuaaano mo oumaaumonoum maouw muaamou umou mo muoumaaumouumvca
saw: moauaaqnm mo oumaaumououm anouw muaommu umou mo muoumfiaumouum>o

 

moawoamwcwwm m mcowumcfinaoo poumcHEocoauuoumumasz

 

 

mHADmMM Emma mo mMOH<ZHHmMIMMQZD mmH MHHS mHADmmm Emma mo
mMOH<EHHmMumm>o osz<mzoo mHADmmM Emma mo mmH¢ZHHmm onB<HmmmmmH2HHmom 92¢ ZOHH<HMMmMMHszmm mcm nz<
mmHHHAHm< mo mMH<EHHmm ZOHH<HMMmmMHzHHmom 924 ZOHafiammmmmHszmm mom NUZ<HM<> ho muz<UHmquHm mo mammauh

>Hxxxxbmam<s

86
of test results than did the over-estimators of test results group,

a difference significant at a highly significant level of confidence.

XVI. SUMMARY OF THE FINDINGS

The major findings of this study are summarized on the following
pages. A considerable number of significant findings were obtained from
the analysis of the data.

1. For the following groups, the post-interpretation means of
estimates of abilities were significantly closer to the points of maximum
accuracy of estimation than the preinterpretation.lmeans: the high and
low academic aptitude groups; and over-estimators and under-estimators
of abilities; and the under-estimators of test results. No significant
changes in means toward the points of maximum accuracy were found for the
total group, the boys, the girls, or the over-estimators of test results.

2. For the following groups, the postinterpretation means of
estimates of test results were significantly closer to the points of
maximum accuracy of estimation than the preinterpretation means: the
high and low academic aptitude groups; the over-estimators and under-
estimators of abilities and the over-estimators and under-estimators of
test results. No significant changes in means toward the points of
maximum accuracy were found for the total group, the boys, or the girls.

3. After test results interpretation there were highly significant
decreases in the variances of the estimates of abilities for the total
group as well as for every sub-group in the study.

4. After test results interpretation there were highly significant

decreases in the variances of the estimates of test results for the total

‘1 I! II. '1‘! I III III Ir.llulllll'|.|li

87
group as well as for every sub-group in the study.

5. No significant differences in accuracy of estimation were
found between preinterpretation estimates of abilities and preinterpretation
estimates of test results.

6. For the following groups, the variances of the postinterpretation
estimates of test results were significantly less than the variances of the
postinterpretation estimates of abilities: the total study group; the
boys and girls, regardless of whether the variances were computed from
the grand means or the appropriate sex mean; the high academic aptitude
group; the under-estimators of abilities; and the under-estimators of
test results. (The differences for the total group, the boys when
computed from the grand means, the girls when computed from the apprOpriate
sex means, the high academic group, the under-estimators of abilities,
and the under-estimators of test results were highly significant.) No
significant differences between the postinterpretation estimates of abilities
and postinterpretation estimates of test results were found for either
the low academic aptitude group, the over-estimators of abilities, or the
over-estimators of test results.

7. When variances of the boys and girls from grand means of estimates
in the study were compared, it was found that there was a tendency toward
less variance in both the preinterpretation estimates of abilities and in
the postinterpretation estimates of test results for the boys, but the
differences did not reach an acceptable level of confidence. Neither were
significant differences in variability found in the preinterpretation

estimates of test results or the postinterpretation estimates of abilities.

 

“I... p15 .. .. a _ littitli

88

8. When the variances from the sex means of estimates of the
boys and girls were compared, it was found that although the boys had
a tendency toward smaller variances in all four estimates, the differences
did not reach an acceptable level of confidence.

9. Neither the total study group nor any of the various sub-
groups were found to differ significantly in the proportion of boys
and girls.

10. When the mean estimates of the high academic aptitude group
were compared with the mean estimates of the total study group, it was
found that all four mean estimates of the high academic aptitude group
were significantly lower than the mean estimates of the total study
group. In addition, the postinterpretation estimates of the high
academic aptitude group were very close to the points of maximum accuracy
of estimation.

11. When the mean estimates of the low academic aptitude group
were compared to the means of the total study group, it was found that
all four mean estimates of the low academic aptitude group were significantly
higher and farther from the points of maximum accuracy of estimation than
were the mean estimates of the total group in the study.

12. When the mean estimates of the over-estimators of abilities
group were compared with the means of the total study group, it was found
that all four estimates of this sub-group were significantly higher and
farther from the points of maximum accuracy than were the mean estimates
of the total study group.

13. When the mean estimates of the under-estimators of abilities

89

group were compared with the means of the total study group, it was
found that all four mean estimates of this sub-group were significantly
lower than were the mean estimates of the total study group. In addition,
the postinterpretation estimates of the under-estimators of abilities
group were very close to the points of maximum accuracy of estimation.

14. When the mean estimates of the over-estimators of test
results were compared with the means of the total study group, it was
found that all four mean estimates of this sub-group were significantly
higher and farther from the points of maximum accuracy than were the
mean estimates of the total study group.

15. When the mean estimates of the under-estimators of test
results group were compared with the means of the total study group,
it was found that all four mean estimates were significantly lower than
were the mean estimates of the total study group. In addition, the
postinterpretation estimates of the under-estimators of test results
group were very close to the points of maximum accuracy of estimation.

16. When the variances of the high academic aptitude group were
compared with the variances of the low academic aptitude group, it was
found that there were no significant differences in the variances of
either the preinterpretation estimates of abilities or of the preinter-
pretation estimates of test results. However, the high academic aptitude
group had significantly less variance in both the postinterpretation
estimates of abilities and the postinterpretation estimates of test results.

17. When the variances of the over-estimators of abilities group
were compared with the variances of the under-estimators of abilities

group, it was found that there were no significant differences in the

9O
variances of either the preinterpretation estimates of abilities or of
the preinterpretation estimates of test results. Although there was
a tendency toward less variance in the postinterpretation estimates of
abilities for the under-estimators of abilities group, the difference
did not reach an acceptable level of confidence. However, the under-
estimators of abilities group had significantly less variance in the
postinterpretation estimates of test results than did the over-estimators
of abilities group.

18. When the variances of the over-estimators of test results
group were compared with the variances of the under-estimators of test
results group, it was found that there were no significant differences
in the variances of either the preinterpretation estimates of abilities
or of the preinterpretation estimates of test results. Although there
was a tendency toward less variance in the postinterpretation estimate
of abilities for the under-estimators of test results group, the difference
did not reach an acceptable level of confidence. However, the under-
estimators of test results group had significantly less variance in the
postinterpretation estimates of test results group had significantly less
variance in the postinterpretation estimates of test results than did the

over-estimators of test results group.

CHAPTER V

SUMMARY, CONCLUSIONS, RECOMMENDATIONS

I. SUMMARY

The purposes of this study were to develOp an approach to the com-
munication and interpretation of test scores, and to use the procedure with
a group of seventh grade pupils in order to investigate its effective-
ness.

The Academic Promise Tests, an aptitude battery measuring abstract
reasoning ability, numerical ability, verbal ability, and language usage
ability, were administered to 281 seventh grade pupils of one urban
junior high school in Indiana. One month later, after a short discussion
of individual differences, abilities, and the concept of norm groups,
the subjects estimated both their abilities in the four areas sampled
by the tests as well as their performance on the tests taken a month
previously. Estimates were made on a chart composed of numbers and
cartoon symbols representing 100 boys and girls arranged in such a manner
as to form a stanine distribution. Immediately after this estimation
procedure, group interpretation of test results followed, using the test
publisher's printed profile. These procedures took place in 10 different
groups averaging 28.1 pupils each. Exactly one week later, the pupils
again estimated their abilities and test results in the same manner as
before.

Data for the study were collected from the 244 pupils (119 girls

and 125 boys) who took the tests and made both preinterpretation and

92

postinterpretation estimates. Means and standard deviations were computed
for the preinterpretation and postinterpretation estimates of both abilities
and test results for the total study group and the following sub-groups:
boys; girls; high academic aptitude pupils; low academic aptitude pupils;
over-estimators of abilities; under-estimators of abilities; over-estimators
of test results; and under-estimators of test results. The t-test technique
was used to test the significance of difference between means, the F-test
technique to test the significance of difference between variances, and the
chi-square technique to test the significance of the prOportions of boys and
girls in the various groups.

The following is a brief summary of the major findings of the study.
After test interpretation, for the total group as well as for every sub-group
in the study, there were highly significant variance decreases in both
estimates of abilities and estimates of test results. No significant difference
in accuracy of estimation were found between the sexes. The interpretation
procedure was especially effective for the high academic aptitude group, the
under-estimators of abilities group, and the under-estimators of test results

group.

II. CONCLUSIONS

From the analysis of the data in this study, the following con-
clusions seem to be justified.

l. The majority of the seventh grade pupils in this study changed
their preinterpretation estimates of their abilities and their preinter-
pretation estimates of their test performance in the desired direction

after participating in the test results interpretation procedure utilized

93
in this study. It is concluded from these data that if seventh grade
pupils similar to the ones in this study receive interpretations of
aptitude test results in a manner similar to the procedure used in
this study, they can be expected to become more accurate in their estimates
of abilities and test performance.

2. While no meaningful differences were found between the pre-
interpretation estimates of abilities and the preinterpretation estimates
of test results, the variances of the postinterpretation estimates of
test results were generally found to be significantly less than the
variances of the postinterpretation estimates of abilities. It is
concluded from these data that junior high school age youngsters are more
able to alter their estimates of their test performance than they are to
alter their estimates of their abilities as a result of group test
interpretation.

3. The eXperimental version of test results interpretation with
groups of pupils was especially effective with the high academic aptitude
group and the groups that under-estimated either their abilities or their
test results. All three of these sub-groups had mean estimates of abilities
and test results very close to the points of maximum accuracy of estimation
as well as highly significant decreases in variability of these estimates
after group test results interpretation. Therefore, it is concluded that
a group test results interpretation procedure similar to the one utilized
in this study is likely to be most effective with high academic aptitude

pupils and those who under-estimate either their abilities or test

performances.

 

94

4. The eXperimental version of test results interpretation with
groups of pupils was somewhat less effective with the low academic aptitude
group and the groups that over-estimated either their abilities or their
test results. However, all three of these sub-groups had highly significant
decreases in variability of estimates of both abilities and of test results
after group test results interpretation. Therefore, it is concluded that
a group test results interpretation procedure shmilar to the one utilized
in this study is likely to be somewhat less effective with low academic
aptitude pupils and those who over-estimate either their abilities or
test performance.

5. No significant differences in performance was found between
the boys and the girls in this study,.and it is concluded that there
is no meaningful difference in the capability of seventh grade boys and
girls to profit from group test results interpretation as practiced in

this study.

III. RECOMMENDATIONS FOR GROUP TEST RESULTS INTERPRETATION

On the basis of the results of this study and the conclusions
formulated above, the following recommendations for group test results
interpretation with junior high school pupils are warranted.

l. The results of this study have several implications for
practical application. It now seems less reasonable to omit inter-
pretation of test results because the available counselor time is in-
sufficient for individual interpretations. For the majority of students,

a group procedure such as the one used in this study is likely to be

95
an effective means of helping a student understand the meaning of his
test scores.

2. It is recommended that before group test results interpre-
tation, the interpreter secure from each pupil an estimate of how he
perceives his relative abilities and how he feels he performed on the
tests. On the basis of this study, it would seem beneficial to follow
up the group interpretation with additional individual or small group
counseling interviews with those students of low academic aptitude and
those who seriously over-estimated either their abilities or their test
performance. This recommendation is in addition to the general invitation
that should be extended to any student who might want to discuss his
test results further with a counselor.

3. The assumption that all pupils understand the concepts of human
variability and individual differences seems unwarranted. It is
recommended that the interpreter preface his interpretation of test
results with a discussion of these concepts.

4. Although this study did not investigate directly the effect
of visual aids, it is the investigator's Opinion that the interpreter make

maximum use of appropriate visual aids during the interpretation process.
IV. SUGGESTIONS FOR FURTHER RESEARCH

The need for further study of the procedures and practices of
group test results interpretation seems apparent. The following are

some suggestions for studies that might very well be undertaken by future

investigators.

”Isl-hi. L .

. l.‘|\ll..r!!' .

t

96

1. As an initial recommendation, this study should be replicated
with other seventh grade pupils as well as with students in other grade
levels. Such replications should be facilitated by the forms and procedures
recorded in the appendix of this dissertation.

2. In this study, the various estimates of the subtests of the
Academic Promise Tests were summed, producing one gross estimation score.
A future investigation might analyze the data further to study possible
differences in estimating the specific abilities and test score estimates.

3. As noted previously, only one week intervened between the date
of interpretation and the follow-up collection of data. A future inves-
tigation might increase or decrease this intervening period in order to
study the effect of time on the ability to integrate and recall such
information.

4. This study used an aptitude test. Other similar studies
might utilize other tests, such as achievement tests, in order to study
any possible differences in estimation accuracy related to the type of
trait being estimated.

5. In a future study, the interpretations of test scores might
be done by the pupils' regular classroom or homeroom teachers. It is
assumed that such interpretations would be preceeded by in-service training
in test interpretation methods and procedures.

6. The contributions of various types of visual aids to the
effectiveness of the interpretation process needs to be studied in detail.
Future research is needed to shed light upon which segments of the inter-
pretation process can be made more effective and efficient by the use of

visual aids, as well as what form these visual aids should take.

97

7. Interpretation groups in this study were composed of pupils
assigned to regular classroom sections of the seventh grade. No attempt
was made to group pupils homogeneously either on the basis of accuracy
of estimations or on the basis of test scores. These factors might be
investigated in order to study the possible effects of the degree of
homogeneity or heterogeneity of the interpretation group on the effective-

ness of the interpretation process.

BIBLIOGRAPHY

BIBLIOGRAPHY

"Are Americans Over-Testing?" Overview, August, 1961,

 

Arnold, Dwight L. "Student Reaction to the Kuder," Personnel and
Guidance Journal, September, 1958. ;

 

Association for Supervision and Curriculum Development, National Education
Association, Perceiving, Behaving, Becoming. Washington, D. C.:
N.E.A., 1962.

Barry, Ruth, and Beverly Wolf, Epitaph for Vocational Guidance. New York:
Bureau of Publications, Teachers College, Columbia University, 1962.

 

Bennett, George K., Marjorie R. Bennett, Dorothy M. Clendenen, Jerome E.
Doppelt, James H. Ricks, Jr., Harold G. Seashore, and Alexander G.
Wesman. Academic Promise Tests Manual, 1962. New York: The
Psychological Corporation, 1962.

 

Berg, Irwin A., "Test Score Interpretation and Client Confusions,"
Personnel and Guidance Journal. May 1956.

 

Bordin, E.S. "A Theory of Vocational Interests as Dynamic Phenomena,"
Educational and Psychological Measurement, 3, 1943.

Combs, Arthur W. "A Perceptual View of the Adequate Personality,"
Perceiving, Behaving, Becoming. washington, D. C.: N.E.A., 1962.

Combs, Arthur and D. Snygg. Individual Behavior, Revised Edition. New York:
Harper & Brothers, 1959.

 

Dressel, Paul L. and Ross W. Matteson. "The Effect of Client Participation
in Test Interpretation," Educational and ngchological Measurement.
x, 4, Winter, 1950.

Durost, W. N. "Present Progress and Needed Improvements in School Evaluation
Programs," Educational and Psychological Measurements, 14.

Erikson, E. H. Childhood and Sociegy. (New York: Norton, 1955).

Friedenberg, Edgar Z. The Vanishinngdolescent. New York: Dell Publishing
Company, 1962.

Froehlich, Clifford P. "Must Counseling Be Individual?" Educational and
Psychological Measurement, XVIII, 4, 1958.

Froehlich, C. P. and W. E. Mesher. "Do Counselees Remember Test Scores?"
Journal 2: Counseling Psychology. 1, Fall, 1954.

99

 

L3"

100

Gibson, Robert L. "Pupil Opinions of High School Guidance Programs,"
Personnel and Guidance Journal, XL, 5, January, 1962.

Goldman, Leo. Using Tests in Counseling. New York: Appleton-Century-
Crofts, 1961.

 

Gustad, John W. and Abdul H. Tuma. "The Effects of Different Methods of
Test Introduction and Interpretation on Client Learning in Counseling,"
Journal g£_Counseling Psychology, Vol. 4, No. 4, 1957.

Hardaway, Charles W., Grace B. Kozak, and Helen Ederle. Budgetagy and
Financial Provisions for Guidance Services in Indiana Public
Secondary Schools. Office of Research and Testing, Indiana State
College, Terre Haute, Indiana, 1963.

 

 

Hastings, g£_gi. "The Use of Test Results - Cooperative Research Project
Number 509? Bureau of Educational Research, University of Illinois,
1960.

Jersild, Arthur T. in Search gf Self. New York: Bureau of Publications,
Teachers College, Columbia University, 1952.

Kamm, Robert B. and Gilbert C. Wrenn. "Client Acceptance of Self-
Information in Counseling," Educational and Psychological
Measurement, X, 1, Spring, 1950.

 

 

Lallas, John E. "A Comparison of Three Methods of Interpretation of
Achievement Tests To Pupils," (unpublished doctoral dissertation,
Sanford University, 1956).

Lane, David. "A Comparison of Two Techniques of Interpreting Test Results
to Clients in Vocational Counseling," (unpublished doctoral dissertation,
Columbia University, 1952).

McCarthy, Dorothea. "Ethical and Professional Considerations in Reporting
Test Information," Proceedings and Summaries, 23d Annual Meeting, New
York State Psychological Association, Inc., May 5-7, 1960.

McNemar, Quinn. Psychological Statistics. New York: John Wiley & Sons,
InCQ’ 1955.

 

1961 Annual Survgngf Standardized Test Publishers, prepared by Stanley B.
Hunt and Associates for the American Textbook Publishers Institute,
432 Park Avenue South, New York City, New York, April, 1962.

Peters, Herman J. and Gail F. Farwell. Guidance: A Developmental Approach.
Chicago: Rand MtNally & Company, 1959.

Phillips, Beeman N. and Garrett Weathers. "Analysis of Errors Made in
Scoring Standardized Tests," Educational and Psychological Measurement,

18, 1958.

lOl

Rogers, Lyle B. "A Comparison of Two Kinds of Test Interpretation
Interview," Journal pf Counseling Psychology, 1954.

Singer, Stanley L. and Buford Stefflre. "Analysis of the Self-Estimate
in the Evaluation of Counseling," Journal pf Counseling Psychology,
I, No. 4, Winter, 1954. A

Super, Donald E. "Vocational Adjustment: Implementing a Self-Concept,"
Occupations, XXX, November, 1951. 7

Super, Donald E. The Psychology pf Careers. New York: Harper & Brothers,
1959.

"Testing: Can Everyone Be Pigeonholed?" Newsweek, 54, 3, July 20, 1959.
"Testing is Big Business," American Psychologist, II, January, 1947.

Thorndike, E. L. "Tests and Their Uses," Teachers College Record, XXVI,
2, October, 1924. .

Thorndike, Robert L. and Elizabeth Hagen. Measurement and Evaluation in
Psychology and Education. Second Edition. New York: Wiley & Sons, 1961.

Tyler, Leona E. The WOrk pf the Counselop, Second Edition. New York:
Appleton-Century-Crofts, 1961.

White, Robert W. The Abnormal Personality. New York: The Ronald Press,
1956.

Wright, Edward Wayne. "A Comparison of Individual and Multiple Counseling
in the Dissemination and Interpretation of Test Data" (unpublished
doctoral dissertation, University of California, 1957).

APPENDIX A

TEST INTERPRETATION SCRIPT

103

Today we are going to spend some time talking about abilities of
seventh graders. Perhaps a good place to start would be to point out that
most of the time we learn about our abilities by comparing our performances
with the performances of others. For example, many seventh grade boys like
to build models, and most seventh grade girls like to dance. You probably
have noticed that before you really know how good you are at building
models or dancing, you have to compare your models or your dancing with the
models and dancing of your friends. As another example, suppose you are
told that on a quiz of 60 spelling words, you got 40 correct. We really
can't say for sure whether or not this is a high score, an average score, or
a low score until we have the answers to at least two questions. First, how
hard are the words? (It would make a big difference if they came from a
college science book or a third grade reader!) Second, what were the scores
of the other seventh graders who took this Quiz? (If you knew this, you
could easily see if your score was among the highest, about in the middle,
or among the lowest.)

This chart1 is supposed to represent lOO seventh grade students of
your age. I would like for you to imagine that one of these symbols re-
presents 123. Therefore, if you are a girl, imagine that this represents
you, 49 girls, and 50 boys, all in the seventh grade. If you are a boy,
imagine that this represents you, 49 other boys, and 50 girls, all in the

seventh grade.

Notice that what we have here is an arrangement of 5 rows of

 

1Appendix B.

U34

20 students each, alternating the boys and girls. This is one way to
arrange our 100 students, but there are many others. For instance, if we
were to arrange them in the order of some ability, we would probably end up
with an arrangement that looks something like this. Here we have the same
100 students arranged in order of their relative amounts of some ability.
Some would demonstrate a high amount of the ability and would rank about here;
some would not have much of the ability and would rank about here; many would
have an average amount of the ability and would rank about here; some would
have an above-average amount of the ability and would rank about here; and
some would have a below-average amount of the ability and would rank about
here. In order that you might see this arrangement a little closer, I have
prepared duplicates on individual sheets, which will be passed out now.2

At the same time we will pass out a mimeographed worksheet.3 Please
print your name and group number on the lines at the top. The schedule
number for this group is ___.

Now place the distribution on top of your worksheet so that the
first paragraph and box shows at the top.

Now, there are many, many abilities that we could discuss today, but
we are going to consider only four. Brief descriptions of these abilities
are printed on your worksheet. Let's take a look at Ability Number 1. Read

the description silently as I read it aloud:

Abstract Reasoning--the ability to see relationships and
recognize concepts presented in the form of diagrams or
symbols rather than in words or numbers.

 

2Appendix B.

3Appendix C.

 

i."i3. II M‘v

. .JIJII;I.:I l
. h . (I

105

Now: look at the distribution. Imagine that these 100 seventh grade
students are arranged from bottom to top in order of their increasing amount
of abstract reasoning ability. That is, those with a low amount of this
ability here, then those with a below-average amount, then those with an
average amount, then those with an above-average amount, and finally, those
with the most of this ability. Remember that this distribution is supposed
to represent lOO seventh graders, including you. Now, in such an arrangement,
find the symbol which you feel best represents the position in this group
that illustrates ygng ability in abstract reasoning, as compared with the
others. (Pause) When you have found that symbol, look to the left and note
the number of the row in which the symbol you have selected appears. Write
that number in the first box on your paper.

OK, slide your distribution sheet down so we can take a look at
Ability Number 2. Read the description silently as I read it aloud:

Numerical--the capacity to think in quantitative terms, to
understand and use numerical relationships.

Look at the distribution again. This time, imagine that these 100
seventh grade students are arranged from bottom to top in order of their
increasing amount of numerical ability. That is, from those with the least
amount of this ability at the bottom on up to those with the greatest amount
at the top. Remember that you are to suppose that one of these 100 students
represents you. In such an arrangement, find the symbol which you feel best
represents the position in this group that illustrates your numerical ability,
as compared with the others. (Pause) When you have located that symbol, look
to the left to see the number of the row it is in. Write this number in the

second box on your paper.

106
Now let's take a look at Ability Number Three on your paper. Read
the description silently as I read it aloud:

Verbal--understanding word meanings and the ability to use
words in reasoning.

This time, imagine that your distribution represents 100 seventh
graders arranged from bottom to top in order of their increasing abstract
reasoningability. As before, those with the least are at the bottom, those
with the most at the top. One of these symbols is supposed to represent
you. Find the symbol which you feel best represents the position in this
group that illustrates your verbal ability, as compared with the others,

Put the number of the row in which your symbol appears in box number three.

Finally, let's consider Ability Number Four on your paper. Read it's
description silently as I read it aloud:

Language Usage--understanding of correct writing and speech,
appreciation of good English.

You probably have guessed correctly that this time your distribution
is supposed to represent 100 seventh graders, including yourself, arranged
from bottom to top in order of an increasing amount of language usage
ability. Find the symbol which you feel best represents the position in
this group that illustrates your language usage ability, as compared to the
others. When you have found it, write the number of the row in which it
appears in box number four. Now cover your worksheet with your printed
distribution.

From this point on, please do not erase or change any number you
have written in the boxes. You will not be graded in any way on this, of
course. The reason is that I am interested in knowing your first impression.

OK?

107

Now, you have just estimated your relative amounts of these abilities.
There are other ways to estimate relative ability. Do you remember this
test you all took about a month ago? (Show.) Well, tests like these also
estimate relative abilities. Do you remember this first part, the pages of
puzzle-like problems? (Show.) Well, this is supposed to be one way of
estimating your abstract reasoning ability.

Look at your copy of this distribution of 100 seventh grade students
again. This time, imagine that these symbols are arranged in order of scores
on this part of the test, from those who made the lowest scores here at the
bottom to those who made the highest scores here at the top. If we were to
arrange 100 seventh grade students who took this abstract reasoning test
in this manner, and you were one of that 100, where do you think you would
be assigned to stand? That is, which symbol do you think represents your
performance on this part of the test as compared to others of your age and
grade who took it? Find that symbol; locate the number of the row in which
it appears, and copy that number on the dotted line next to the first box.
'Npgg: this number may or may not be the same as the one you have previously
'written in the box. If it is, OK; if it is larger, OK; if it is smaller,
OK; this is up to you to decide.

Do you remember the second part of the test, the one with the
arithmetic problems? This part of the test is supposed to estimate your
numerical ability. Imagine this time that the 100 students in the dis-
tribution are arranged from bottom to t0p in order of scores on this part
of the test, from those who made the lowest scores here at the bottom to
those who made the highest scores here at the top. If we were to arrange

lOOIseventh grade students who took this test in this manner, you were

108

one of that 100, where do you think you would be assigned to stand?
Which symbol do you think represents your performance on this numerical part
of the test as compared to others of your age and grade who took it? After
you have located that symbol, see what row it is in, and c0py that number
on the dotted line beside box number two. As before, this number may or may
not be the same as the one you put in the box previously.

Remember the part of the test that was concerned with word meanings,
and how words were related to each other? This part of the test attempts
to estimate verbal ability. Imagine that your distribution now represents
100 seventh grade pupils arranged in order of their scores on the verbal
part of the test, from those who made the lowest scores at the bottom to
those who made the highest scores at the top. If you were one of the 100
students arranged in such a manner, where do you think you would be assigned
to stand? Which symbol do you think represents your performance on this verbal
part of'the test as compared to others of your age and grade who took it?
Locate that symbol; note the number of the row in which it appears, and write
that number on the dotted line beside the third box. As before, this number
may or may not be the same as the one you wrote on the box number three.

Finally, do you remember the last part of the test in which you were
asked to inspect some sentences that were divided into three parts and
tell whether there was an error of some kind in part A, part B, part C, or
if there were no errors? This part of the test is supposed to estimate language
usage ability. Look at your distribution and imagine this time that the 100
students are arranged from bottom to top in order of their increasing scores
on the language usage part of the test. That is, with those who made the lowest

scores at the bottom, and those who made the highest scores at the top. If

109
you were one of the 100 students arranged in this manner, where do you think
you would be assigned to stand? Which symbol do you think represents your
performance on this verbal part of the test, as compared to others of your
age and grade who took it? Find your symbol; note the number of the row
in which it appears, and write that number on the dotted line beside the
last box. As before, this number may or may not be the same as the one you
wrote in box number four.

Check your worksheet; you should have your name printed at the top,
your group number, ____, in the prOper space, one number in each of the
boxes, and one number on each of the dotted lines. Now pass your work-
sheets to the front of the room and then the printed distributions also.

Now let's review what we have done so far. You have estimated your
relative abilities in abstract reasoning, numerical, verbal, and language
usage. Then you estimated your performance on the tests that measures these
abilities. The next step will be to reveal to you your test results. Please
keep your test profile sheet face down until everyone has theirs. Also,
although you may not care if others see your test results, perhaps they
might not want you to see theirs. For this reason, it will be best if you
keep your eyes on your own paper.

Turn your papers over. The printing on the left describes what this
test is supposed to measure, and how to use this profile sheet in under-
standing your scores. We are interested in the first four columns. Let's
look at the first column, which tells us about your results on the abstract
reasoning part of the test. Notice the two numbers above this first column.
The one in the gray area does not concern us now, but the one in the white

area, beside the word "percentile" does. A percentile tells what per cent

 

 

110
of the students of your age and grade who took this part of the test made
scores 35 or pgipg your score. Think of it as being the number of students
out of 100 who made scores at or below yours. This number says to you,
"Out of 100 students of your age and grade who took this abstract reasoning
part of the test, your score is as high or higher than this many."

Now look at the column below the first percentile number. Note the
heavy black line in the middle, the one with the number 50 beside it. A
score at this point would have approximately 50%.or half of the other scores
above it, and 502 or half of the other scores below it. Part of your column
has been blackened with a crayon; this blackened area indicates how far and
in what direction your score varied from the middle of the distribution.

If your percentile score at the top of the first column is greater than

50, the blackened area should run up from the 50 point to your score.

If your score was less than 50, the blackened column should run down to your
score. Checkb see that the percentile number at the tOp is the same as the
number at the end of the blackened column.

The second column reports your performance on the numerical part of
the test. Find the percentile number at the top. As before a percentile
tells what per cent made scores at or below your score. Think of it as being
the number of students out of 100 who made scores like or below yours. This
number says to you, "Out of 100 students of your age and grade who took this
numerical part of the test, your score is as high or higher than this many."

Check the blackened area of the second column to see if it has been
marked correctly. If your percentile score is larger than 50, the column
should run up from the 50 point to your score. If your percentile score is

less than 50, the column should run down from the 50 point. The number at

111
the end of this column should be the same as the percentile number at
the top.

Column three shows your performance on the verbal section of the
test. Find the percentile number at the top of the column, which is inter-
preted just like the other percentiles: it shows what per cent of other
seventh grade students who took this part of the test made scores at or
below your score. It indicates the number of students out of 100 who made
scores like or below yours. Think of it as saying to you, "Out of 100
students of your age and grade who took the verbal part of the test, your
score is as high or higher than this many."

Check the blackened area of this column to see if it has been colored
in correctly. If your score is larger than 50, the column should run up from
the 50 point to your score. If your score is less than 50, the column should
run down from the 50 point to your score. As before, the number at the end
of this blackened column should match your percentile score at the top.

Finally, column 4 indicates your performance on the language usage
section of the test. The percentile number at the top shows what per cent
of the other students like you made scores at or below your score. It can
be thought of as the number of students out of 100 who made scores like or
below yours. It says, "out of 100 students of your grade who took this
language part of the test, your score is as high or higher than this many."

Check the column 4 blackened area to see if it has been marked
correctly. If your percentile score is larger than 50, it should run up
from the 50 point. If your score is less than 50, it should run down from
the 50 point. In either case the number at the side of the end of this

column should be the same as the percentile number at the top.

112

One thing further might be said about percentile scores. In
general, scores that fall between 30 and 70 can be said to be about
"average," those between 70 and 85 "above average," those between
15 and 30 "below average," those above 85 "high,' and those below
15 "low." These are rough approximations, but are close enough for
our purposes here.

Now who has a question I can try to answer?

In just a minute or so, we will take up your profiles. In a
few weeks I will return them to Mr. Dobson in the Guidance Office on
the second floor. Perhaps you would like to arrange an individual
interview to discuss your test results. If so, he has asked me to
tell you that he will be very happy to make an appointment with any

student who would like to have one.

APPENDIX B

NORM GROUP CHART IN STANINE FORM

 

114

3:
“.22:
32:23:

3:233:33;

2222:2223:
@322;ng
223222
@332

:2

a” N GO In H' M N H

APPENDIX C

ESTIMATION WORKSHEET

116

Your name: Group number:

 

(Last) (Middle) (First)

Ability No. 1: Abstract Reasoning

 

 

and recognize concepts presented in
the form of diagrams or symbols
rather than in words or numbers ..........

...the ability to see relationships I

 

 

Ability No. 2: Numerical

 

 

quantitative terms, to understand

...the capacity to think in A
and use numerical relationships

 

0.0.0.0...

 

Ability No. 3. Verbal

 

the ability to use words in reason-

...understanding word meanings and i ]
ing

 

Abilipero. 4. ,Language Usage

 

...understanding of correct writing
and speech, appreciation of good
English

——b

 

 

_E

 

oooooooooo

sssssssss

----------

APPENDIX D

FOLLOW-UP SCRIPT

118

Today we would like for you to take a few minutes to make a few
more estimates for us. First we will pass out the materials, the printed
distribution chart1 and a worksheet.2 (Pause) Print your name and group
number in the indicated spaces at the top of the worksheet. The schedule-
number of this group is‘_____.

Perhaps you were absent on the day of the test a month ago, or
perhaps you were absent a week ago when I was here before. If so, write,
"I was absent on the day of the test," or "I was absent a week ago" above
your name on the worksheet.

One caution: please do not jump to conclusions as to what we will
be doing today! These directions are a little bit different than they
were last week. Listen carefully until each section of the directions
has been completed so you won't make a mistake and have to erase.

Let me say at this point that no one in this school will see your
responses on this worksheet, so I hope you will try to be as accurate in
your estimations as possible. And of course you will be in no way graded
on your responses; I am just interested in learning what you think.
Because this is a matter between you and me, it will be best if you keep
your paper covered.

Let's review briefly what this distribution is supposed to represent.
Remember that we pointed out that if we arranged 100 individuals in order
of their increasing amounts of some ability, we would probably get a

distribution similar to this one. Those with the least amount of the ability

 

1Appendix B.

2Appendix C.

119
would rank about here, those with a little more about here, those with an
average amount in this region, those with an above-average amount about here,
and those with the most of the ability here. For convenience the rows have
been numbered.

The first thing we would like for you to do today is to indicate
your understanding of your test scores. First, place the distribution
sheet on t0p of your worksheet in such a manner that all that shows of the
worksheet are the dotted lines on the right and the group number at the
top.

Look at the distribution. Imagine that these are 100 seventh grade
students arranged from bottom to top in order of their scores on the first
section of the test you took, the abstract reasoning section. One of these
symbols is supposed to represent you. Find the symbol that is located in
the position that indicates your test score in abstract reasoning as compared
to the others. Look to the left, note the number of the row in which this

symbol appears, and write that number on the first dotted line.

 

Next, imagine that these 100 seventh grade students are arranged
from bottom to top in order of their scores on the second section of the
test you took, the numerical section. As before, one of the symbols is
supposed to represent you. Which symbol is standing in a position that
represents your numerical test score as compared to the others? Look to
the left, note the number of the row in which this symbol appears and
write that number on the second dotted line.

Now, imagine that these 100 seventh grade students are arranged from
bottom to top in order of their scores on the third section of the test,

the verbal section. One of them is supposed to be you. Which symbol is

120
located in the position that indicates your test score on the verbal
section of the test as compared to the others? When you have found that
symbol, look to the left, note the number of the row in which it appears,
and write that number on the third dotted line.

Finally, imagine now that the 100 seventh grade students are
arranged from bottom to top in order of their scores on the last part
of the test, the language usage section. As usual, one of these symbols
is supposed to represent you. Find the symbol that is located in the position
that indicates your test score in language usage as compared to the others.
Look to the left, note number of the row in which your symbol appears, and
write that number on the last dotted line.

Now, cover your worksheet with your distribution sheet.

Please pay careful attention to the next directions. Remember
that we pointed out last time that tests are only one way of estimating
abilities. There are many other ways, including making your own estimates.
This is what we would like for you to do next, to indicate your own estimations
of your relative abilities.

Place the distribution sheet on t0p of the worksheet so that the
dotted lines and the group number space is covered. On the left you should
see the descriptions of four abilities, each followed by a box.

Read the description of the first ability silently as I read it
aloud:

Abstract Reasoning--the ability to see the relationships and

recognize concepts presented in the form of diagrams or symbols
rather than in words or numbers.
Look at the distribution. Imagine that you see 100 seventh grade students

arranged from bottom to t0p in order of increasing abstract reasoning

121

ability. One of them represents you. How much abstract reasoning ability
do you think you have in comparison with the others? Find the symbol which
is located in a position that you feel accurately indicates your abstract
reasoning ability as compared with the others in the distribution. (Pause)
Note the number of the row in which it appears, and write that number in
the first box. Please note: this number Esy_or may 925 be the same as the
one you wrote on the first dotted line. If you feel that it should be
larger, OK, if you feel it should be smaller, OK, and if you feel that it
should be the same, OK. We are interested in knowing what you think, how
you estimate your abstract reasoning ability.

Read the description of the second ability silently as I read it
aloud:

Numerica1--the capacity to think in quantitative terms, to
understand and use numerical relationships.

Look at your distribution. Imagine that you see 100 seventh grade students
arranged from bottom to tOp in order of increasing numerical ability. One
of them represents you. How much numerical ability do you think you have
in comparison with the others? Find the symbol which is located in a position
that you feel accurately indicates your numerical ability as compared with
the others in the distribution. (Pause) Note the number of the row in which
it appears, and write that number in the second box. As before, it may or
may not be the same as the number you wrote on the second dotted line. We
are interested in knowing what you think, how you estimate your numerical
ability.

Finally, read the description of the third ability as I read it

aloud:

122

Verbal-~understanding word meanings and the ability to use
words in reasoning.

Look at the distribution, and this time imagine that the 100 seventh
grade students are arranged from bottom to t0p in order of increasing verbal
ability, one of them representing you. How much verbal ability do you think
you have in comparison with the others? Find the symbol which is located
in a position that you feel accurately indicates your numerical ability as
compared with the others in the distribution. (Pause) Note the number of
the row in which it appears, and write that number in the third box. As
before, it may or may not be the same as the number you wrote on the third
dotted line. We are interested in knowing what you think, how you estimate
your verbal ability.

Finally, read the description of the fourth ability silently as I
read it aloud:

Language Usage--understanding of correct writing and speech,
appreciation of good English.

Look at your distribution, and imagine that the 100 seventh grade students
are arranged from bottom to top in order of increasing language usage
ability, one of them representing you. How much verbal ability do you
think you have in comparison with the others? Find the symbol which is
located in a position that you feel accurately indicates your numerical
ability as compared with the others in the distribution. (Pause) Note the
number in the last box. As usual, it may or may not be the same number as
you wrote on the last dotted line. We are interested in knowing what you
think, how you estimate your language ability.

Take a look at your paper--you should have your name and group

number, , at the top, a number from 1 to 9 on each of the dotted

123
lines, and a number from 1 to 9 in each of the boxes.
Pass your worksheets to the front of the room, face down. Then

pass in the distribution sheets.

‘mmm“