Group Title: investigation of the Cannot Say scale of the group Minnesota multiphasic personality inventory
Title: An investigation of the Cannot Say scale of the group Minnesota multiphasic personality inventory
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00097966/00001
 Material Information
Title: An investigation of the Cannot Say scale of the group Minnesota multiphasic personality inventory
Alternate Title: Minnesota multiphasic personality inventory
Physical Description: vi, 68, 1 leaves. : illus. ; 28 cm.
Language: English
Creator: Eaddy, Morris Lee, 1935-
Publication Date: 1962
Copyright Date: 1962
 Subjects
Subject: Personality tests   ( lcsh )
Psychology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis--University of Florida, 1962.
Bibliography: Bibliography: leaves 39-41.
Additional Physical Form: Also available on World Wide Web
General Note: Manuscript copy.
General Note: Vita.
 Record Information
Bibliographic ID: UF00097966
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000546107
oclc - 13174134
notis - ACX0065

Downloads

This item has the following downloads:

PDF ( 3 MBs ) ( PDF )


Full Text











AN INVESTIGATION OF THE CANNOT SAY

SCALE OF THE GROUP MINNESOTA

MULTIPHASIC PERSONALITY

INVENTORY








By
MORRIS LEE EADDY


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY









UNIVERSITY OF FLORIDA
February, 1962















ACKNOWLBDGEMENTS


The writer is indebted to many people who contrib-

uted to the completion of this study* Dr. Dorothy

Rethlingshafer, the Chairman of the writer's committee,

gave much of her time to the early formulation of the

study's proposal and helped the writer along the way with

her constructive criticisms and encouragement.

Two members of the writer's committee were replaced

by new members before the completion of the study.

Dr. Justin Harlow died suddenly of a heart attack on

October 3, 1961. Dr. Malcolm Robertson left his position

on the staff to accept a new appointment elsewhere*

Dr. James Dixon and Dr. Audrey Schumacher consented to

fill their vacancies and the writer is grateful for

their assistance* All the members of the committee,

Dr. Benjamin Barger, Dr. James Dixon, Dr. Justin Harlew,

Dr. Milan KolarikO Dr. Dorothy Rethlingshafer, Dr. Malcolm

Robertson, Dr. Audrey Sohunacherv and Dr. Bruce Thomason .

are thanked for the aid they freely gave to the writer*

Special mention must be made of Dr. Benjamin Barger

who consulted on many occasions with the writer regarding

the use of the MMPI and who gave many valuable suggestions*

MMPI data were made available by Dr. Barger*

-ii-








The facilities for obtaining data from the Strong

Vocational Interest Inventory were made available through

the University Counseling Center, directed by Dr. Justin

Harlowe The writer appreciates deeply the assistance

given him by Dr. Harlow, the cooperation of the secretarial

staff of the University Counseling Center, and the per-

mission to conduct moat of the testing in apace made

available by the University Counseling Center*

The writer very much appreciates the help given his

on the statistical treatment of data by Dr. Harry J.

Wahler of the Columbus Mental Hygiene Clinic, Columbus,

Ohio* Other aid was given by Dr. Vincent O'Connell of

the Columbus Psychiatric Institute and Hospital

In any study of this sort, the subject who gave

their time and energy to make the study possible deserve

the writer's full thanks and appreciation.

Finally, the writer expresses his appreciation to

his wife, Norma Lee, for her constant support and

encouragement, her patience and interest expressed through-

out the months involved in the completion of the study.


*iiil















TABLE OF CONTENTS


ACKNOWLEDGEMENTS . . . .

LIST OF TABLES * * * * .

CHAPTER

I. INTRODUCTION . . .

II. EXPERIMENTAL DESIGN .

III. RESULTS . . . o

IV* DISCUSSION . . . .

V, SUMMARY AND CONCLUSIONS

BIBLIOGRAPHY . . . .

APPENDICES

APPENDIX A. STATISTICAL TABL

APPENDIX B. TEST INSTRUMENTS


* 0 0 0 0 0 .




* 0 0 0 0 0
. 0 0 0 0 0 0 0




. . . . .


* 0

* 0


E


* 0

* 0


S 0 0 0 0


APPENDIX C.


NUMBER AND PERCENTAGE OF UNIVER-
SITY OF FLORIDA FRESHMEN WHO


WERE HIGH CANNOT SAY SCORERS
GIVING CANNOT SAY RESPONSES TO
EACH ITEM 0 * * * * *

BIOGRAPHICAL SKETCH *. . . * o s


-IV-


Page

ii











23

29

36

39




43

49















LIST OF TABLES


Table Page

1. STRONG TOTAL NON-COMMITTAL SCORES OBTAINED
FROM THREE GROUPS OF SUBJECTS, MALE AND
FEMALE, SELECTED ON THE BASIS OF MMPI ITEM
OMISSION BEHAVIOR . . . . . . a

2. ANALYSIS OF VARIANCE OF MALE TOTAL NON-
COMMITTAL SCORES OBTAINED FROM THE STRONG
FOR THREE GROUPS OF SUBJECTS SELECTED ON
THE BASIS OF MMPI ITEM OMISSION BEHAVIOR .* 4

3. ANALYSIS OF VARIANCE OF FEMALE TOTAL NON-
COMMITTAL SCORES OBTAINED FROM THE STRONG
FOR THREE GROUPS OF SUBJECTS SELECTED ON
THE BASIS OF MMPI ITEM OMISSION BEHAVIOR o. k

h. MEAN DIFFERENCES BETWEEN STRONG TOTAL
NON-COMMITTAL SCORES FOR THREE GROUPS OF
MALE SUBJECTS SELECTED ON THE BASIS OF
MMPI ITEM OMISSION BEHAVIOR . . . o h4

5. STRONG "?" SCORES OBTAINED FROM THREE
GROUPS OF SUBJECTS, MALE AND FEMALE, SE-
LECTED ON THE BASIS OF MMPI ITEM OMISSION
BEHAVIOR . . . * * * * * t5

6. ANALYSIS OF VARIANCE OF MALE "?" SCORES OB-
TAINED FROM THE STRONG FOR THREE GROUPS OF
SUBJECTS SELECTED ON THE BASIS OF MMPI
ITEM OMISSION BEHAVIOR . . . . . 45

7. MEAN DIFFERENCES BETWEEN STRONG "?" SCORES
FOR THREE GROUPS OF MALE SUBJECTS SELECTED
ON THE BASIS OF MMPI ITEM OMISSION BE-
HAVIOR . . . . . . . . . h6

8. ANALYSIS OF VARIANCE OF FEMALE "?" SCORES
OBTAINED FROM THE STRONG FOR THREE GROUPS OF
SUBJECTS SELECTED ON THE BASIS OF MMPI ITEM
OMISSION BEHAVIOR * 1 o a o o h46


-*v









Table


9. MEAN DIFFERENCES BETWEEN STRONG "?" SCORES
FOR THREE GROUPS OF FEMALE SUBJECTS SELECTED
ON THE BASIS OF MMPI ITEM OMISSION BE-
HAVIOR a 0 * 0 0 0 , 0 a h7

10* ANALYSIS OF VARIANCE OF INTOLERANCE OF
AMBIGUITY, IN TERMS OF MALE AND FEMALE CET
SCORES, FOR THREE GROUPS OF SUBJECTS SELECTED
ON THE BASIS OF MMPI ITEM OMISSION BEHAVIOR. 47

11. ANALYSIS OF VARIANCE OF DESIRE FOR CERTAINTY,
IN TERMS OF MALE AND FEMALE DC SCORES, FOR
THREE GROUPS OF SUBJECTS SELECTED ON THE
BASIS OF MMPI ITEM OMISSION BEHAVIOR . . h8

12. TEST OF THE SIGNIFICANCE OF THE DIFFERENCE
BETWEEN THE MEAN NUMBER OF MMPI OMISSIONS FOR
TWO GROUPS OF MMPI ITEMS CATEGORIZED ON THE
BASIS OF SOCIAL DESIRABILITY RATINGS . . 8


-vi-


Page















CHAiiPTE'i I


INTROtC1 T'I



itat
Since itt. publicztior in 194Z2, the Minnesota Multi-

phesic Personality Invoutory (IKPI) has undergone extensive

investigation and now stands as our most widely used per-

sonality test. Cronbach (1960) states that it has been

studied more extensively and more adequately than any test

of its kind which probably accounts in large measure for

its present general acceptance and wide usage.

The MMPI is composed of ten "clinical scales" and four

"validity" scales. The "clinical scales" (Hs, D, Hy, Pd,

Mf, Pa, Pt, Sc, Ma, and Si) have been investigated to a

considerable extent. Of the "validity" measures, however,

only three--L, F, and K--have received much attention from

research workers. The fourth "validity" measure, the

Cannot Say scale, which consists simply of the counted num-

ber of items omitted on the test, has thus far received

relatively little consideration in the literature.

A large number of Cannot Say responses is considered

undesirable in an MMPI protocol and various means are

employed to either (a) reduce the number of such responses


-1-









-2-

by giving special instructions to the test subject or (b)

correct for depressed profiles resulting from excessive

item omissions by certain methods of adjusting the scores

(Brown, 1950; Hovey, 1958)o Some investigators (Dahlstrom

and Welsh, 1960) recommend eliminating altogether the

Cannot Say alternative.

The effect of a high number of omissions in a MMPI

protocol is commented upon by Dahlstrom and Welsh (1960):

"Any item placed in the Cannot Say category is
automatically removed from the scoring of the
MMPI scales. It is not considered as either
deviant or conforming, True or False. The general
effect of such omissions, then, is to reduce the
length of the test, shrink the variance, and
attenuate the profile."

Due to the omission of a large number of items the

examiner is unable to make comparisons with the standard-

ization populations. Though there may be a question

whether or not a claim for lessened validity is in all

cases true, it would seem safe to assume that the

reliability of the test is reduced because of the con-

sequent reduction in the length of the test* These authors

recommend the wording of instructions so as to influence

test subjects to respond to every item in the test.

However, since item omission itself is test behavior

the possibility arises that when excessive omissions occur

in a protocol, even after instructions are worded so as to

minimize their occurrence, important data might be wasted

by discarding the test. The Cannot Say score might provide








d3-

information concerning personality characteristics of the

test subject.



Purpose

The purpose of this study is (a) to investigate

whether excessive item omission behavior on the MMPI may

contribute "clinical" data by reflecting certain

personality characteristics of the test subject and (b)

to consider certain characteristics of the test itself

as possibly influencing item omission behavior.



Survey of the Literature

That valuable information may lie in the high Cannot

Say score is suggested by Hathaway and McKinley (1951)

who states

"In its own right the Question score is an indi-
cator of personality factors, but no specific
clinical material on it has been analysed."

Hanley (1957) also mentions this possibility#

"...reference to outside criteria may reveal,
as has often been suggested, that the validating
scales themselves tap personality dimensions rather
than temporary response sets."

In his study of control in psychological adjustment,

Caudra (1956) eliminated from consideration any

protocol which contained more than 40 omitted items.

He states

"The fairly large number of records which were re-
jected*..suggests that the height of the Cannot Say
score may be a good diagnostic clue; this was not
pursued further, however, being only tangential to
the purpose of the present research."








-2&.

Tankin and Scherer (1957) recognized this possibility

and attempted, although unsuccessfully, to relate certain

personality characteristics to high Cannot Say scorers.

In line with the probability that such test-taking

behavior may reflect personality variables, the question

arises "How stable is item omission behavior?" There

is evidence available which indicates that normal subjects

tend to be fairly consistent in the number of items

omitted when retested with the NMPI up to intervals of

1,060 days between tests (Hoelberg and Alessi, 1949;

Kaufmann, 1956; Schofield, 1950, 1953). This suggests

that item omission behavior may be a fairly stable

characteristic and, as Hanley (1957) contends, not merely

a temporary response set. Guilford (1954), using the

Guilford personality inventories STDCR, GAMIN, and

Personnel Inventory found considerable stability in the

tendency of certain factory foremen to respond with "?"

from test to test. To date, however, there has boon no

attempt reported in the literature to relate the con-

sistency of item omission behavior on the MMPI to

similar behavior on other tests. This study investigates

the consistency between item omission behavior on the

MMPI and certain similar test behavior on the Strong

Vocational Interest Inventory. It a consistency is found

to exist, support will be given to the hypothesis that

the tendency of individuals to omit items on the MMPI is

not "test-specifie" but is a more general personality








-5-

trait or behavior characteristic.

That the tendency of some individuals to respond to

test items in certain ways regardless of item content

is a relatively stable personality trait was suggested by

Cronbach (1946) some years ago. He was one of the first

to concern himself with the problem of "response sets"

as variables which affect psychological test performance.

He contended that "response sets" such as acquiescence,

tendency to gamble, tendency to disagree, and tendency

to use the neutral response in "...tests of attitudes,

personality and psychophysical judgments," are character-

istic of certain individuals. He suggested that the

unstructured or ambiguous nature of some test items is

probably the most important factor influencing the appear-

ance of such "response sets". Berg and Rapaport (1951)

found evidence to support the belief that "response sets"

occur when test subjects are tested in an ambiguous

situation. Hamilton (1957) cites data obtained from

testing neurotic and control subjects in a variety of

ambiguous situations and states*

"The data show marked individual differences and
considerable evidence for intra-individual con-
sistency in avoidance/non-avoidance of ambiguous
situations."

He suggests that the neurotic person may avoid am-

biguity in order to reduce the anxiety which occurs as a

result of the uncertainty and conflict aroused by such

situations. That is, individuals were found to have








-6-

different degrees of intolerance of ambiguity, a construct

which has been dealt with thoroughly by Frenkol-Brunsvik

(1948s 1949).

A concept which appears to be closely related to

intolerance of ambiguity is Brim's concept of desire for

certainty (1955)e Interested in attitude research data

which pointed to the fact that some individuals respond

more intensely to attitude items than other individuals,

he attempts an explanation of this behavior as an

expression of the individual's desire or need for cer-

tainty. He states

"Our explanation of individual differences in
generalized intensity in response to attitude
items is that people high in intensity have a
greater need for security, and so respond with
greater convictionuo.an individual high on in-
tensity must also make more frequent responses
at the extreme of the...content dimension, be-
cause only in this way can he actually remove
the ambiguity."1

These two constructs, intolerance of ambiguity and

desire for certainty, offer the possibility of con-

tributing to a better understanding of the test item

omission behavior observed in some individuals who take

the MMPI. The items which compose the MMPI vary in the

specificity of the behavior to which they refer. That is,





lBrim and Hoff (1957) state that further studies
dealing with the need for security indicate that a more
accurate description of what Brim's test measures is a
desire for certainty. They therefore use the term
"certainty" rather than security in speaking of this
test.









-7-

some items may be considered to be more ambiguous than

other items. Also, the MMPI, because of the rather

intimate personal content of many of the items, may be

reacted to as a somewhat threatening situation by some

individuals. The relative ambiguity of the individual

items on the MMPI, along with the possible threat2 of

the test itself, may be factors which influence test

performance of some individuals.

One way of expressing intolerance of ambiguity in

response to an ambiguous test item is for the individual

to "leave the field", that is, to not commit himself

by omitting the item. The lack of structure of the

item, the lack of certainty the individual feels with

regard to the item, may be resolved by not marking the

item either True or False. This study seeks to determine

whether the need for structure, for certainty (which seem

to be closely related constructs), influences some

individuals to omit items on the MMPI.



ZThat items on the MMPI may pose something of a
threat to some test subjects is also suggested by a
personal communication from Hovey (1960). Responding to
this investigator's request for further information re*
garding his method of correcting for excessive item
omissions (1958), he commented briefly concerning his
interest in an investigation along the lines of the
present study
"I did not continue a comparable study here
because of difficulty in obtaining additional cases for
cross-validation* It seems that word passed around among
patients here that it is not at all dangerous to answer
all the items* At any rate, an answer sheet seldom shows
up with as many as 25 items left unanswered."









-8-


While the major interest of this study is to inves-

tigate personality characteristics of individuals who

omit a large number of items on the MMPI, data which are

readily available make it possible to investigate one

characteristic of the test itself which may influence

item omission. Edwards (1959), using Cronbach's (1949)

observations of response sets as a starting point, gives

evidence to support the hypothesis that, "The social

desirability of items on a personality test markedly

influences the probability of their being endorsed."

More specifically stated his hypothesis is<

"...just as individual differences have been found
in the tendencies of subjects to respond True, Un-
decided, or False, regardless of item content, so also
there are individual differences in the tendencies
of subjects to give socially desirable responses
to items in personality inventories, regardless of
whether the socially desirable response is True or
False."

Heineman (1960) asked 108 students in Introductory

Psychology to rate each item of the MMPI on a five-point

rating scale of favorability. He reports favorability

ratings for each of the 566 items. These data make it

possible to test the hypothesis regarding the effect of

social desirability of the items on item omission be-

havior. As has been suggested by Hanley (1957) and

Heineman (1953), the social desirability content of MMPI

items may influence some test subjects to respond in terms

of putting themselves in a good light rather than in

terms of giving an honest opinion of themselves. In fact,







-9-


many writers feel that the degree of social desirability

of item content acts as a major factor affecting responses

to personality measures* This study will test the

hypothesis that some test subjects may be influenced in

their responses to the MMPI items by the social desire

ability of the items and omit largely those items which

may be said to reflect socially undesirable character-

istics.



Statement of Hypotheses

The following hypotheses are investigated in this

study

*1 Subjects who omit a high number of items on the MMPI
will exhibit similar test-taking behavior on the
Strong Vocational Interest Inventory.

2. Subjects who omit a high number of items on the MMPI
will show less tolerance of ambiguity than subjects
whose number of omissions falls at the mode or around
the mean for the entire population under consideration.

3. Subjects who omit a high number of items on the MMPI
will be found to have a greater desire for certainty
than subjects whose number of omissions falls at the
mode or around the mean for the entire population
under considerations

h. Subjects who omit a high number of items on the MMPI
will omit a significantly larger number of items which
may be considered to reflect socially undesirable
characteristics, as opposed to items on the MMPI which
may be considered to reflect socially desirable
characteristics.
















CHAPTER II


EXPERIMENTAL DESIGN



Sample

The subjects for this experiment consisted of second

year male and female undergraduate students at the

University of Florida who completed the MMPI and the

Strong Vocational Interest Inventory as part of the re-

quired testing program of the University prior to their

Freshman year in college* Three different groups were

selected*

High Cannot Say Group (BCS)o Subjects composing the

HCS group were individuals whose Cannot Say score was

equal to or greater than approximately two standard

deviations from the mean number of omissions for the entire

population of 3,198 students tested in September of 1959.

The omission score of 24 was thus selected as the cut-off

point for the ECS group*

Mid-Range Cannot Say Group (MCS). A second group con-

sisted of subjects whose number of omissions on the MMPI

fell in the neighborhood of the mean number of omissions

for the entire population. This group included students


-10-







-11-

whose number of item emissions was from four to six.

Low Cannot Say Group (LCS)o This group was composed of

subjects whose Cannot Say score fell at the mode of the

distribution* The mode of the Cannot Say distribution

was found to be sereo



Test Instruments

McReynolds' Concept Evaluation Technique (CET) (1951, 1950).

The CET, as modified by Eriksen (1953), was used as the

measure of intolerance of ambiguity in this study. It

consists of 50 Rorschach concepts, half of which are

scored plus and half minus by Beck's frequency tables*

This subject is asked to accept or reject (Yes or No)

each of the 50 Rorschach concepts indicated by the ex-

aminere

McReynolds' criteria for the selection of the con-

cepts is specified as follows

"The selection of concepts from Beck's list was
made in such a manner that for each Rorschach card
there was the same or nearly the same number of
plus concepts and minus concepts, i.e*, for each
plus concept with given typical determinants (move-
ment, form, color, and the like) there was a cor-
responding minus concept; and for each plus concept
represented by a given size of blot area (W,D,DdS)
there was a corresponding minus concept* Of the 50
concepts, 5h per cent were from the achromatic
cards and 34 per cent were from the last three
cards" (1951).

McReynolds found the reliability of the number of

Yes'es (i*e*. acceptance of suggested concepts) to be *82

for a total of 214 cases of mixed psychiatric diagnosis,









-12-

as estimated by the odd-even method and corrected by the

Spearman-Brown formula. More pertinent to the present

study, he obtained a reliability coefficient of .93 for

a group of 69 normal subjects* The percentages of Yes

responses to each item were also computed for normal men

and women to see whether separate norms would be needed,

but no significant differences between the two groups

were found (1954).

While McReynolds did not make use of the number of

rejections as a measure of intolerance of ambiguity in his

studies he did comment on possible interpretations of the

J-score (his term for the number of Yes'es recorded for a

test subject). He states$

"More specifically, the hypothesis for a low J-score
would be that this individual tends to maintain, in
his thinking, strict standards of evaluation, to
have highly channelised thought patterns which would
tend to revolve endlessly about a few subjects" (19514)

The number of rejections has been used by several other

investigators as a measure of intolerance of ambiguity.

Briksen (1953), studying personality rigidity and the

Rorschach, found that$

"Subjects who reject a large number of indicated
Rorschach concepts or interpretations also tend
to show a greater reluctance to offer hypotheses
to an ambiguous stimulus and less ability to free
their behavior from perceptual-cognitive sets."

He found that the number of rejections on the CBT corre-

lated -.47 with "hypothesis availability" when subjects

were confronted with visual stimuli tachistoscopically








-13-


presented at sub-rocognition threshold speeds This

negative correlation was significant at the .01 level.

A negative correlation of -.*2, significant also at

the .01 level, was found between the number of rejections

and the number of direct solutions on problems 6 through

11 of the water-jar problems described by Luhbins.

A correlation of .40 was found to exist between the

number of rejections on the CIT and the offset of an

erroneous oxpoetanoy upon speed of perceptual recognition.

Davids (1955) also used the CBT as a measure of

intolerance of ambiguity but failed to show any signif-

ioant relationships between performance on the CRT and

measures of authoritarianism.

The findings of Kenny and Ginsberg (1958), however,

show a lack of correlation among a battery of tests which

purport to measure intolerance of ambiguity.

"Only 7 of the 66 correlations among measures of
intolerance of ambiguity were significant at the
.05 level, two of those having a relationship
opposite to that predicted."

This lack of correlation suggests that caution must be

used in employing any test purporting to measure this

construct. They further states

"In so far as the present measures may be regarded
as relevant indicators of the intolerance of ambiguity
construct, the results offer little support for a
general construct of intolerance of ambiguity***.These
conclusions do not mean, however, that the construct
should necessarily be discarded. It may, rather, be
less general or broad in scope than had been initially
assumed* Future research may discover a number of
distinct or relatively independent dimensions of
intolerance of ambiguity rather than just one unique
generalized factor."








*14-

The specificity of intolerance of ambiguity measures,

as suggested by Kenny and Ginsberg, indicates the need

for critical evaluation of the test behavior of subjects

before attaching any moaning to their test performances.

One obvious meaning of such test behavior is that some

people are simply more suggestible than others. McReynolds

entertained this possibility but pointed to a study done

by lysenek and Furneaux (1945) which, he felt, indicated

that, "***suggestibility as indicated, e.g., by hypno-

tisability, is not significantly correlated with the

willingness of S's to accept suggested concepts on ink-

blots" (McReynolds, 1951), Sinee other possible inter-

pretations could also be offered to explain test behavior

on the CET, the rationale for using the number of re-

jections of the suggested Rorschach concepts as a measure

of intolerance of ambiguity deserves consideration*

The reasoning has been of this natures The person

vho is intolerant of ambiguity will require more structure

than the person who is tolerant of ambiguity before

accepting suggested concepts of the ambiguous visual

stimuli. Half of the concepts are "good form" responses

and half are not* Thus, the person who is intolerant

of ambiguity is expected to reject a larger number of

concepts than the person who is more tolerant of

ambiguity since half the concepts lack "good form" quality.

However, an argument may logically be made for a

contrasting point of view* The person who is intolerant











of ambiguity supposedly will try to impose structure on

ambiguous situations in order to remove the uncertainty

implied in this sort of situation. On the CLT, the

examiner goes so far as to suggest possible interpre-

tations of indicated torschach concepts, thus giving

structure to the different parts of the blots. The person

who is intolerant of ambiguity, because of his need for

structure, might be expected to accept more concepts than

the person who is tolerant of ambiguity.

Thus, there exists the possibility of interpreting

responses to the CET from either of two points of view*

It appears reasonable, however, to assume that subjects

who are intolerant of ambiguity, even though possible

interpretations of the ink blot areas are offered by the

examiner, would still maintain "stricter standards of

evaluation" than subjects who are tolerant of ambiguity.

No doubt numerous interpretations could be made to each

of the concepts but the test subject must respond to the

one concept suggested. McReynolds (1954) likens this

procedure to a paired comparison technique in that's

"In taking the CET the S is required to compare a
blot area with a suggested concept. For each
evaluation then, the S is presented with two concepts-
that implied by the question to him (e.g., bat, man)
and that of the blot area. His task is to compare
the two concepts and determine whether they are
alike (Yes answer) or different (No answer)* The
procedure may therefore be considered a paired
comparison technique. If the S's standards of com-
parison are strict he will tend to answer No| whereas
if he answers Yes more frequently it is indicated
that his standards are less strict."









-16-


This study, like that of Eriksen (1953) and Davids

(1955) makes use of the total number of rejections of

Rorsehach concepts on the CBT as a measure of the striet-

ness of the subject's standards of evaluation, or rather

his intolerance of ambiguity.

Desire for Certainty Test (DC). This test, termed a

protective measure by its author, purports to measure,

"...individual differences in the strength of the desire

for certainty" (Brim and Hoft, 1957). The test consists

of 32 statements about everyday events in the following

forms "The chances that an American citizen will believe

in God are about in 100." Subjects respond by filling

in a probability estimate for each statement and then

indicating how certain they feel about their estimate by

rating it on a five-point scale ranging from 1 (Very Sure)

to 4 (Not Sure at All). The author states$

"The statements...(include)...four from each of
eight different areas, e.g., economiose recreation.
Half of the statements (are) in conformity to
generally accepted American values, while half (are)
not. In addition each statement had a different es-
timated probability value drawn from a set of 32
values ranging from 1 to 46 and from 54 to 99.
Combinations of these three criteria were made
randomly as specifications for the construction of
each test statement" (Brim and Hoff, 1957).

Brim attempted to control for the amount of knowledge

subjects had about an item by "...drawing the items from

a variety of areas." The use of a homogeneous group,

such as college Freshmen and Sophomores, minimized

differences in the amount of general information possessed








*17-

by the subjects* In the construction of the test the

author selected toot items front different content and

value areas in order to randomiae the effects of iter

desirability upon toot performance* No relationship was

found to exist between subjects' performance on the DC

teot and standard socioeoonomic variables or with

intelligence

Regarding the rationale underlying the DC test, Brim

states$

*The relationship between probability estimates of
oeourrenco of events and the certainty of these
estimates show the typical U-shaped curve found in
the content-intensity relationship of attitude
data. An explanation in terms of an equiprobability
through ignorance principle is given (here) for both
the attitude and probability expectation data. In-
dividual differences in both intensity and extroeity
of responses to attitude and expectancy questions are
conoidored to be the result of individual differences
in need for security. A stronger need results in
proportionately more items being responded to with
greater intensity and extroeity as the individual
seeks to avoid the middle range of response in which
he admits hip ignorance" (Brim, 1955).

Brim feols that the construct of desire for certainty is

closely akin to the oonstruot of intolerance of ambiguity*

Using an odd-oven split, corrected by the Spearman-

Brown formula, the reliability of this toot was found to

be .81 on an N of 50 subjects (Brim and Hoff, 1957).

Correlations with the personality ratings of

roomaates reported by Brim and Hoff were found to be .18

and *24 respectively when separate scoring methods for

(a) extroeity of probability estimate and (b) indicated

certainty of response were used* Hoeveor, when a joint









-18-

scoring method was used (i.e., difference of probability

estimate from 0 or 100, multiplied by the indicated

certainty or probability estimate) a correlation of *39

was obtained which was significant at the .01 level*

Brim (1955) provides evidence to support the belief that

a joint scoring method is a more reasonable way of ob-

taining an estimate of the subjects' desire for certainty*

Thirty-eight subjects filled out six attitude state-

ments from different areas and indicated their agreement

or disagreement on a 13-point continuum An average

extremity score was computed for each subject* Extremity

scores on the attitude were correlated with scores on the

desire for certainty test* An r of *57 was obtained which

was found to be significant at the *01 level (Brim and

Hoff, 1957).

Brim and Hoff tested the hypothesis that, "***high

and low scorers on the F scale would be similar in showing

a high desire for certainty." Using eta as a test for

curvilinear relation a value of *28 was obtained, which

has a probability of between *05 and .10 (two-tailed)

both for significance and for nonlinearity.

The average extremity scores obtained from 83 sub-

jects' responses to three Thurstone scales on censorship,

patriotism, and the Bible were also correlated with scores

on the desire for certainty test* By computing a Pearson

product-moment coefficient of correlation an r of *28,









-19-

significant at the .02 level, was obtained by Brim and

:off.

They conclude from this data that:

"Significant correlations were obtained between
response extremity and test scores, suggesting
that individual differences in desire for
certainty are consistent from one measure to
another."

In another study reported in the same article, they

report that they were able to experimentally increase or

decrease the degree of desire for certainty exhibited by

subjects on the DC test.

The present study used the joint scoring method

recommended by Brim. Responses were scored for the tend-

ency to make estimates approaching 0 or 100, and the

degree of certainty for the estimates indicated. Thus,

if a subject responded to a particular item with a

probability estimate of 10 out of 100 and indicated that

he was Very Sure (i.e., circled "1") of his estimate,

his joint score for the item was 10. The joint scores

for all 32 items for each subject were summed and an

average score obtained. Hence, the lower the mean score

on this test, the higher the desire for certainty of the

subject.

Strong Vocational Interest Inventory. This well-known

interest inventory was used to assess the degree of

generality which can be claimed for test item omission

behavior of subjects who take the MMPI. If this test-

taking behavior is not merely a temporary "response set"








-20-

it should occur to some degree on other tests.

Certain responses on the Strong are similar to test

item omission behavior on the MMPI. Items 1-280 of the

Strong for men, items 1-255 and items 362-400 for women,

may be responded to by the marking of one of three possible

alternatives Like, Indifferent, and Dislike. The

"Indifferent" response is believed to be similar in nature

to item omission in the MMPI in that the subject does not

have to reveal his feelings. It is a Non-Committal

response*

Another Non-Committal alternative appears in the

Strong with the "Equally Well" response, items 321-360

for men and items 296-333 for women. Here the subject

indicates his choice of pairs of activities by respond-

ing with Like, Dislike, or Like Both Equally Well. The

"Equally Well" response is considered to be similar to

item omission on the HMPIO

One other response category on the Strong which is

similar to item omission on the MMPI is the "?" response,

items 366-388 for men and items 362-400 for women.

Thus, by counting the number of "Indifferent",

"Equally Well", and "?" responses on the Strong, a Total

Non-Committal score was obtained for each test subject.



Procedure

Thirty subjects, 15 males and 15 females, from each

of three groups, formed on the basis of MMPI item omission









-21-


behavior: (a) a high Cannot Say group (HCS); (b) a mid-

range Cannot Say group (MCS)l (o) a low Cannot Say group

(LCS) were tested with the CET and the DC test* Pre-

ceding the administration of the CET the subject was

given a brief description of the Rorschach boards and

asked to look at each card quickly and tell some of the

things he saw or that might have been represented on the

cards. Following MeReynolds' recommendations for the

administration of the CRET the procedure below was

utiliteds

E. "Have you ever taken or do you know anything
about the Rorschach, the ink blot test?"

S, (Responds to question.)

B. "Well, this is a test which consists of ten cards
which have on them designs made up out of ink
blots. Just to familiarize you with them, look
at each of them quickly and tell me some of the
things you see or some of the things which
might be represented." (e hands the cards, one
at a time, to the 3 and waits for S to respond.
After one or two responses* n takes the card and
hands S the next one.)

g. (After the last card has been seen by 8) "So far
we have gone through the cards and you have told
me the things that you have seen (and you have
done very well)* Now we are going through them
quickly again, and this time I am going to suggest
certain things to you. I want you to tell me it
you can see them. That's all there is to it this
time. I will say* 'Could this be a--something or
other--O and you look at it, decide and say, 'Yes,
it could' or 'Noe it couldn't.' Now, seoe of the
things I will suggest to you look like what I
say, and some don't--so dent say anything could
be what I say unless you really see it( on the
other hand, be sure to tell me if you do see it*
Don't take too long to study each thing I
suggest, but decide as quickly as you can whether
it could be what I suggest. Do you understand?








-22-


All right, here*s the first one" (McReynolds,
1951).

After the subject responded to each of the 50 con-

cepts with either a Yes or No response, he was immediately

presented with the DC test. The following instructions

were given for this tests

"Indicate on the answer sheet provided you
the probability (some number from 0 to 100)
you feel is appropriate to the following
statements* Indicate also, by circling one
of the numbers (1, 2, 3, 4, or 5) the degree
of certainty you feel with regard to your
probability estimate. Do not spend too much
time on any one statement and finish as
quickly as you can. Reemeber, each probability
estimate refers to so many chances out of 100."

After reading the instructions the examiner asked the

subject if there were any questions and explained again

the instructions if needed.

For 64 males and 34 females in the HCS group, 69

males and 40 females in the MCS group, and 56 males and

38 females in the LCS group, Strong answer sheets were

obtained. The number of "Indifferent", "Equally Well",

and "?" responses on the Strong were counted for sub-

jects in each of the three groups and Total Non-Committal

scores obtained. The "?" scores were also listed

separately.

For the HCS group alone (Male N m 661 Female N 36)

the total number of item omissions for each of the 566

MMPI items was tallied*















CHAPTER III


RESULTS



To test this study*' first hypothesis that subjects

who omit a high number of items on the MKPI will exhibit

similar test-taking behavior on the trong Vocattional

Interest Inventory, the Total Non-Committal scores de-

scribed earlier were computer for males and female. in

each of the tree groups under consideration. Since the

items making up the 'tron_ differ in content for males

and females, anu since the number of items included in

each of the three sub-divisions of the Total Non-Committal

score varied ("Indifferent", "Equally Well", and "?"),

the interest inventory for males and females could not be

considered identical, and the data obtained for males and

females were treated separately (Table 1).

A simple analysis of variance for groups of unequal

sizes was run using the computational formulas outlined

by McNemar (1955). For the three male groups an F of

4)32, significant at the .05 level, was obtained in-

dicating that the degree of MMPI test item omission be-

havior is a real source of variation. That is, when the

Total Non-Committal Strong scores for the three male

-23-








*24-

groups were considered simultaneously, the three group

means were found to differ significantly among them-

selves (Table 2). The degree of item omission on the

MMPI was found to be related to similar test behavior on

the Strong.

Using the t test for independent samples the

differences between the means of all combinations of the

three groups were tested for significance. A t of 2o31,

significant at the .05 level, was obtained when the

mean Total Non-Committal score of the HCS group was com-

pared with the mean Total Non-Committal score of the LCS

group. When the mean of the HCS group was compared with

the mean of the MCS group, a t of 2.61, significant at

the *01 level, was obtained. Comparison of the mean of

the LCS group with the mean of the MCS group, however,

produced a difference which was not significant, giving

a t of .22 (Table 4). Thus, for the male group the first

hypothesis is supported.

Utilizing the Total Non-Committal scores obtained

on the Strong by the three female groups a non-significant

F (F .79) was obtained (Table 3).

Since the "?" items of the Strong could be considered

the most similar in nature to the MMPI Cannot Say alter-

native, these data were considered also and, as before,

separately for males and females. A simple analysis of

variance for groups of unequal sixes resulted in an F of

4.32, significant at the .05 level when the three male








-25-


groups were considered (Table 6). The mean Strong "?"

score for the BCS group was found to be significantly

greater than the mean score of the LCS group at the *05

level) similarly the BCS Strong "?" group mean was

greater than the Strong "?" mean of the MCS group at the

*02 level* No significant difference was obtained when

the difference between the mean Strong "7" scores for the

LCS and the MCS group were compared (Table 7).

Looking at the data obtained from the three female

groups, however, some interesting differences are noted.

A simple analysis of variance for groups of unequal

sizes yielded an F of 7.38, significant at beyond the .01

level (Table 8). Comparing the mean Strong "?" score of

the HCS group with the mean Strong "?" score of the LCS

group, the mean of the BCS group was found to be signif-

icantly greater at the *001 level. In contrast to the

results obtained with the male subjects, however, there

was no significant difference found between the means of

the HCS group and the MCS group (t a .38); also, in con-

trast with the results obtained for the male subjects, the

female MCS group obtained a mean Strong "?" score which

was significantly greater, at the *01 level, than the

mean score obtained by the LCS group (Table 9). Thus,

for the females, the HCS group and the MCS group were

quite similar in their responses to the Strong "?" items

and both significantly different from the performance

of the LCS group.








-26-

The study' second hypothesis stated that subjects

who omit a high number of items on the nMMP will show

greater intolerance of ambiguity than subjects whose

number of omissions falls at the mode (sero omissions)

or around the mean (tour to six omissions). Using an

odd-even split corrected by the Spearian-Brovw formula the

reliability of the CBT was found to be .83 for the total

N of 91 subjects tested in this study. This is slightly

lover than the reliability estimate of .93 reported by

MoReynolds (1954).

Scores on the CET for subjects in the three groups

were determined and a 2-way classification analysis of

variance was run. Rove columan and interaction effects

were not found to be significant. Neither sex, degree of

MMPI item omission behavior, nor an interaction of these

were significantly related to the subjects' rejection of

Rorschach concepts on the CET (Table 10).

The third hypothesis concerned the relative degree

of desire for certainty among the three groups of sub-

jects. Using an odd-even split corrected by the

Spearman-Brovn formula, the reliability of the DC test

was found to be .90 for the total N of 92 subjects tested.

This is somewhat higher than the reliability estimate of

.81 reported by Brim and Heff (1957).

Again a 2-way classification analysis of variance

was used, sex and group membership being considered. Rov,

column, and interaction effects were not significant,









-27-

though the effect of the row variance (i.e., sex) did

approach significance (Table 11).

It is seen then that subjects who responded with a

high number of omissions on the MMPI were not found to

exhibit a greater intolerance of ambiguity, as measured

by the CET9 and did not show a greater desire for cer-

tainty, as measured by the DC test, than the subjects

who omitted fever MMPI items (zere er four to six

omissions)*

The last hypothesis concerned the influence of social

desirability of the items composing the MMPI upon the

likelihood of subjects omitting certain items* It was

hypothesized that subjects who emit a high number of

items on the MMPI tend to omit items which may be con-

sidered to refer to socially undesirable characteristics.

The frequency of omissions for each item on the MMPI of

subjects in the RCS group was tabulated. Each item was

divided into one of two classifications (a) Favorable

and (b) Unfavorable. Since Heineman's five-point rating

scale on favorability was used all items included in the

range of .00 to 3*00 were classified as Favorable and all

items included in the range of 3901 to 5*0 were classified

as Unfavorable* On this basis, 221 of the items were

classed as Favorable and 345 as Unfavorable. Means and

standard deviations were computed and a t test computed to

determine the significance of the difference between the








-28-

two group means. The difference between the two means

reached significance at the .05 level but in the

direction opposite that predicted (Table 12). Subjects

were found to have a tendency to omit items which were

supposed to reflect desirable characteristics.














CHAPTER IV


DISCUSSION



The results obtained from the Strong uay be inter-

preted as providing some support for the hypothesis that

item omission type behavior is relatively stable from one

test to another, although sex and the specific measures

used to compare test behaviors are important variables*

When the Total Non-Committal response was used as the

measure of behavior similar to the MMPI Cannot Say re-

sponse, item omission type behavior was found to be

relatively stable for sales but not for females.

However, when the Strong "?" score alone was used

as the measure, the number of omissions on the MMPI was

found to be a significant factor for both males and fe-

males, though still with some differences as noted

earlier. That is, whereas for the males the BCS group

obtained significantly higher Strong "?" scores than

either the MCS or the LCS groups, for females the HCS

group and the MCS group did not differ significantly

from each other but did differ significantly from the

LCS group*

It appears then that some support is present for


-29-








-30-

considering MMPI test item omission behavior as a general

test taking behavior characteristic, and not merely test

specific, though sex differences clearly emerge and com-

plicate the picture. If the Strong "?" response may be

considered to be more closely akin to the Cannot Say

response of the MMPI than is the Total Non-Committal

score, the sex difference becomes less of a consideration

though still in evidence.

If the reasons for an individual omitting items on

a questionnaire are considered numerous possibilities are

immediately apparent. Dahlstrom and Welsh (1960) gives

the most thorough review of possible reasons behind ex-

cessive use of the Cannot Say category of the MMPI. Con-

fusion, highly charged emotional reactions to the test

items, indecisiveness, dysphoric mood of the depressed

patient, guardedness and/or suspiciousness, rationalizing

the non-applicability of certain items (i.e., "motivated

errors"), and lack of cooperation are some of the

possibilities mentioned. Brown (1950) distinguishes

between justifiable and unjustifiable omissions. Justi-

fiable omissions include items which actually do not

apply to the individual or items about which the in-

dividual really does not know enough to answer either

"True" or "False". Unjustifiable omissions include

deliberate attempts to falsify a response (i.e., respond

Cannot Say) when in reality the individual could answer

"True" or "False", and omissions due to indecisiveness.








-31-

He states that the Cannot Say alternative, ".,.provides a

protective cage where the examine can retire at will when

he feels threatened," Hathaway and McKinley (1951) mention

that some MMPI users have observed high item omissions

by patients with psychasthenia and retarded depression.

Tankin and Scherer (1957) investigated the hypotheses,

drawn from statements of both Brown (1950) and Hathaway

and McKinley (1951), that's (a) the Cannot Say scale,

"...reflects a defensive or evasive attitude on the part

of the examinee...", and (b) it is, "...related to the

symptoms of psychasthenia and depression in psychiatric

patients." Using MMPI profiles obtained from 12 male

psychiatric patients whose Cannot Say scores ranged from

26 to 145 they were unable to substantiate either of these

two hypotheses. They felt that, in general, "High Cannot

Say scores..edo not seem to represent a defensive attitude

on the part of the psychiatric patient," though they

added that certain individuals might use the Cannot Say

alternative in this manner. They recognized that their

investigation, however, was limited to MMPI variables

alone and state that their findings do, "...not preclude

the possibility that other meaningful relationships may

exist, which were outside the scope of this study."

The present study, like that done by Tamkin and

Scherer, selected only several of many possibilities to

test. Using subjects selected exclusively from a college

population, and using variables other than MMPI performance,








-32-

this study found no significant difference in measures

of intolerance of ambiguity or desire for certainty in

subjects who were high emitters on the MMPI as compared

with subjects who were low omitters. These results should

be examined from several aspects.

The most obvious question concerns the tests selected

to measure intolerance of ambiguity and desire for cer-

tainty. As mentioned earlier, Kenny and Ginsberg (1958),

using a variety of tests which purport to measure intoler-

ance of ambiguity, found a lack of support for a general

construct of intolerance of ambiguity* Though the CBT

was not one of the measures used and thus did not enter

directly into the results they obtained, their findings

suggest caution in terming any one test a measure of

intolerance of ambiguity* Studies utilizing the CIT as

a measure of this construct have been few and its validity

may be questioned on the grounds of inadequate research*

This objection could certainly be raised, however, con-

cerning any of the measures which have been used in the

past to assess this construct* It is this investigator's

belief that the validating work which has been done with

the CBT (Eriksen, 1953) furnishes a reasonable degree of

evidence regarding its validity. Assuming that the DC

test is similar to the CBT in the behavioral character-

istics it measures, the same may be said to justify its

inclusion in this study.

Assuming then a reasonable degree of validity for the








-33-


CBT and the DC test, it must be concluded that neither

intolerance of ambiguity or desire for certainty, two re-

lated constructs, were factors which significantly in-

fluenced subjects' use of the Cannot Say alternative.

Bria and Hoff believe that the degree of an

individual's desire for certainty partly reflects that

individual's lack of a feeling of security. To go a bit

beyond this, one might say that the individual who mani-

tests a desire for a high degree of certainty is reflect-

ing a lack of self-sufficiency, a lack of a feeling of

confidence in his ability to meet and deal adequately

with his environment. Rubin-Rabson (1954) tested k3

adults, 18 to 58 years of age, with several tests in-

eluding the Hunter Test of Social Attitudes and the

Bernreuter Personality Inventory and found a negative

correlation of -.*9, significant at the .01 level, between

the Bernreuter measure of self-sufficiency and the number

of "non-committal" responses on the Hunter Test. She

states, "A feeling of self-inrvfficiency is a basic

dynamism in the avoidance pattern, whether of persons,

activities, or iders." Ae~umini that the DC test measures

to some degree an individual's degree of self-sufficiency,

the negative results obtained in the present study appear

to contrast with the finding of Rubin-Rabson. If the DC

test does measure a behavior characteristic similar to

self-insufficiency, the HCS group would be expected to

obtain significantly lower scores (indicating a greater










desire for certainty and, therefore, less self-sufficiency)

than the MCS or the LCS groups. This, however, was not

the case as the number of omissions on the MMPI did not

prove to be a significant factor in the DC scores obtained

by the three groups.

The results obtained by Rubin-Rabson and those ob-

tained by the present investigator are not directly

comparable since different tests were used and since the

populations under consideration were dissimilar in several

respects. Though the above comments are speculative, it

would be interesting to obtain measures of self-sufficiency

for subjects in groups similar to the groups which wore

used in this study and determine whether or not they

differ significantly in the degree of self-sufficiency

evidenced. wolft (1955) has developed from an internal

consistency analysis a self-sufficiency scale (Sf) uti-

lising 3 MMPI items. If future research is found to give

added support to the use of the St scale as an index of

general subjective certainty it will be possible to make

use of Sf scores to investigate the hypothesis that sub-

eocts who are high emitters on the MMPI evidence less

self-sufficiency than subjects who emit fewer MMPI items.

Such a study could also provide information concerning

the appropriateness of considering the DC test as a re-

lated measure of self-sufficiency.

It is difficult to account for the results obtained

regarding the social desirability hypothesis. The HCS










-35-


group tended to omit items which may be considered to

reflect socially desirable personality characteristics*

This may perhaps suggest a general tendency in these

subjects to maintain a sore critical self-concept or one

which reflects a lack of self-acceptance. There are, of

course, other possibilities also*

The data utilized in testing the social desirability

hypothesis of necessity were restricted to the HCS group

since the low number of emissions in the other two groups

did not permit their inclusion* Thuse there was no way

to tell whether or not subjects who omit fever items on

the MMPI also tend to omit items which reflect socially

desirable characteristics* The results speak only for

the HCS group.

Future research will perhaps shed more light on the

clinicall" significance of the Cannot Say seale* The

present study, together with the study of Taakin and

Scherer, has failed to confirm the validity of several

hypothesized personality correlates which were thought

to be involved in excessive item omission behavior. It

remains the writer's belief, however, that the Cannot

Say scale does tap certain personality characteristics

which may provide valuable data to the clinician.

Further research on this scale would appear to be warranted

and needed.















CHAPTER V


SUMMARY AND CONCLUSIONS



Using University of Florida Sophomores, 30 subjects,

15 males and 15 females, from each of three groups formed

on the basis of MHPI item omission behavior were tested

with McReynolds' Concept Evaluation Technique (CET) and

Brim's Desire for Certainty test (DC). The CET was in-

cluded as a measure of intolerance of ambiguity and the

DC test was used to measure a similar construct, need

for certainty. The three groups tested were (a) a

High Cannot Say group (HCS) including subjects whose

Cannot Say score was equal to or greater than 241 (b)

a Mid-Range Cannot Say group (MCS) including subjects

whose Cannot Say score was from four to six{ and (c) a

Low Cannot Say group (LCS) including subjects whose

Cannot Say score was zero*

Total Non-Committal scores were obtained for three

similar groups of larger size by summing the "Indifferent",

"Equally Well", and "?" responses on the Strong Vocational

Interest Inventory*

For the EC8 group alone the number of item emissions

for each of the MMPI items were counted and each item


-36-









-37-

placed in one of two olassifioatiensa (a) Socially

Desirable or (b) Socially Undesirable*

It was found that the number of item emissions on

the MMPI was not significantly related to the scores ob-

tained by subjects on either the CKT or the DC test.

The stability of item omission behavior vhen MMPI

scores were compared with similar behavior on the Strong

was found to vary with the sex of the subjects and the

specific Strong measure used. Using a Total Non-Committal

score as the Strong measure, males who were high emitters

on the RMPI were found to have significantly greater

scores on the Strong. No significant difference was ob-

tained when the female groups were considered. When the

Strong "?" score alone vas considered, however, the

influence of the number of item omissions on the MMPI was

found to be significant for both males and females,

although the female MCS group was not found to differ

significantly from the female HCS group.

The HCS group tended to omit items which may be con-

sidered to reflect favorable personality characteristics

to a significantly greater extent than items which may be

considered to reflect unfavorable personality character-

istics.

From the findings presented in this study it seems

reasonable to conclude the followings

*1 Intolerance of ambiguity and desire tor certainty,
as measured by the CBT and the DC test respectively, are










not significant factors in influencing item omission be-
havior on the MMPI.

2. Stability of item omission behavior on the MMPI,
as determined by relating the consistency of this behavior
on the MMPI to similar behavior on the Strong, depends in
part on the sex of the individuals concerned and in part
upon the specific measure used to compare test behaviors.
Some degree of stability can be inferred but the generality
of item omission behavior still required further inves-
tigation.

35 Subjects who omit an excessive number of items
on the MMPI omit a significantly greater number of items
which may be said to reflect socially undesirable per-
sonality characteristics* This general trend may be re-
lated to certain aspects of the subjects' self-concept,
perhaps a lack of self-acceptance.

4i The writer believes that the Cannot Say scale of
the MMPI may contain information regarding personality
characteristics of the test subject though as yet there
has been little evidence brought forth to substantiate
this possibility.


-38-














BIBLIOGRAPHY


1. Berg, I. A., and Rapaport, Go MN Response bias in
an unstructured questionnaire. Jo Psychol.,
195o, 38, 8 75-481.

2. Bri, 0O. G., Jr. Attitude content-intensity and
probability expectations. Amer. Social.
Review, 1955, 20, 68-76.

3o and Hoff, Do B. Individual and situational
differences in desire for certainty. Jo
abnora* soo. Psychol., 19579 5, 225-229.

h. Brown, MN No Evaluating and scoring the Minnesota
Multiphasic "Cannot Say" items. J olin.
Psychol., 1950, 6, 180-184.

5. Caudra, M. No A scale for control in psychological
adjustment (Cn). In Welsh, Go H., and Dahlstroa,
W. G, (Bds.). Basic readings on the MMPI in
pstyhology and iedicineo NinneapoTist nTiversity
of Minnesota Press, 1956.

6. Cronbaoh, L. Jo Response sets and test validity*
Edu.o psychol. Measmt., 1949, 6,e 75-9,b.

7. essentialss of psychological testing.
(2nd Ed.) New Yorks Harper and Brothers, 1960.

8. Dahlstrome W. Go0 and Welsh, G. S. An MMPI handbooks
a guide to use in clinical practice and
research" Minneapolis: University of
Minnesota Press, 1960.

9. Davids, A. Some personality and intellectual
correlates of intolerance of ambiguity. Jo
abnorm, sooe Psychol., 1955, 51, 415-420.

10. Edwards, A* Lo Social desirability and personality
test construction. In Bass, B. M., and Berge
I. A. (Eds.). Objective approaches to per-
sonality assessment* Nei Yorks D. Van
Nostrand Co.a, Inc, 1959.


-39-









-40-


11 Erikeen, C. W., and Eisenstein, D. Personality
rigidity and the Rorschach. J. Pors., 1953,
21, 386-391.

12. Nysenok, H. J., and Furneaux, W. D. Primary and
secondary suggestibility an experimental and
statistical study. J. exp. Psychol., 1945,
35, 485-503.

13. Frenkel-Brunsvik, N. Intolerance of ambiguity as
an emotional and perceptual variable. J. Pore.,
1949, 18, 108-143.

i1. Tolerance toward ambiguity as a personality
variable. Amer. Psychologist, 1948, 3, Z68
(Abstract).

15. Guilford, J. P. The validation of an "indecision"
score for prediction of proficiency of foremen.
J. appl. Psychol., 195k, 38, z22-226.

16. Hamilton, V. Perceptual and personality dynamics
in reactions to ambiguity. Brit. J. Psychol.,
1957, 48, 200-215.

17. Hanley, C. Deriving a measure of test-taking
defensiveness. J. consult. Psychol., 1957,
21, 391-397.

18. Hathavay, S. R., and McKinley, J. C. Minnesota
Multiphasie Personality Inventoryi Manual
(Rev.). New Yorks Psychological Corporation,
1951.

19. Hoineman, C. 3. A forced-choice form of the Taylor
anxiety scale. J. consult* Psyohol., 19539
17, 447-454.

20. Favorability rating of each item (of the MMPI).
In Dahlstroem VW Go, and Welsh, G. S. (Eds.)
An MNPI handbooks a guide to use in clinical
practice and research. Minneapolist University
of Minnesota Press, 1960, 430-433.

21. Holsberg, J. D., and Alessi, 8S Reliability of the
shortened MMPI. J. consult. Psyohol., 1949,
13, 288-292.










041-


22. Hoveyr, H B. Correction of MMPI profiles for ex-
cessive item omissions. Unpublished materials,
1958. Reported in Dahlstrom, W. 0., and
Welsh, G. 8. (Edso) An MMPI handbooks a guide
to use in clinical practice and research.
Minneapolisa University of Minnosota Press,
1960.

23. Personal communications* 1960.

24, Kaufmann, P. Changes in the MMPI as a function of
psychiatric therapy. In Welsh, 0. 8., and
Dahlstrom, WV G., (Eds*) Basic readings on
the MMPI in psychology and Redicineo
Minnoepoliss Universit' of Minnesota Press,
1956, 525-533.

25. Konny, D. T., and Ginsberg, R. The specificity of
intolerance of ambiguity measures. J. abnorm.
soc. Psychol., 1958, 56, a00-304.

26. MeNemar, Qo Psychological statistics (2nd Ed.).
Nev Yorkt John Wiley and ons, Inc., 1955.

27. McReynolds, P. Perception of Rorsohach concepts
as related to personality deviations. J.
abnorms soo Psychol., 1951, 46, 131-14T.

28, The Roroshach concept evaluation techniques
7 proj. tech., 1954 18, 60-74.

29. Rubin-Rabson, QG Correlates of the non-oommittal
test-item response* J. olin. Psychol., 1954,
10, 93-95.

30. Schofield, V. A further study of the effects of
therapies on MMPI responses. J. abnoer. soe.
Psychol., 1953, 48, 67-77.

31. Changes in responses to the MPIZ following
certain therapies. Psychol. Monogro, 1950,
64, No. 5 (Whole No. 311).

32. Tamkin, A* S, and Scherer, I. VW What is measured
by the "cannot say" scale of the group MMPI?
J. consult. Psychol., 1957, 21, 370-371.

33. Wolff, W. M. Certaintys Generality and relation
to manifest anxiety. Jo abnorm. soc. Psychol.,
1955, 50, 59-64.





































APPBNDIC3S














APPENDIX A


STATISTICAL TABLES



TABLE 1

STRONG TOTAL NON-COMMITTAL SCORES OBTAINED
FROM THREE GROUPS OF SUBJECTS, MALE AND
FEMALE, SELECTED ON THE BASIS OF MMPI
ITEM OMISSION BEHAVIOR


SCS MCS LCS
N M SD N M SD N M SD

Males 64 103.06 36.31 69 126.83 33.55 56 128.16 33.27

Females 34 117.68 30*33 40 110.58 38*23 38 107.32 35.51
i _11 i i I II 1 I I ] : : : . .. : :


TABLE 2

ANALYSIS OF VARIANCE OF MALE TOTAL NON-COMMITTAL SCORES
OBTAINED FROM THE STRONG FOR THREE GROUPS OF SUBJECTS
SELECTED ON THE BASIS OF MMPI ITEM OMISSION BEHAVIOR



Source of Variation SS df Variance Estimate

Between Groups 10io06 2 5,203.00

Within Groups 224z010 186 1,204.35

Total 234,416 188
... .. .. . . . ,_ I . _ _


F 4.32 (Significant at .05 level)


-43m











TABLE 3


ANALYSIS OF VARIANCE OF FEMALE TOTAL NON-COMMITTAL SCORES
OBTAINED FROM THE STRONG FOR THREE GROUPS OF SUBJECTS
SELECTED ON THE BASIS OF MMPI ITEM OMISSION BEHAVIOR



Source of Variation 88 df Variance Estimate


Between Groups 1,995 2 997.5

Within Groups l37,439 109 1,260.9

Total 139,434 111


F .79 (Not Significant)



TABLE 4

MEAN DIFFERENCES BETWEEN STRONG TOTAL NON-COMMITTAL SCORES
FOR THREE GROUPS OF MALE SUBJECTS SELECTED ON THE BASIS
OF MMPI ITEM OMISSION BEHAVIOR



roups NMI t a

CS vs LCS 1k.90 2.31 *

CS vs MCS 16.23 2.61 **

LCS S MCS 1.33 .22


a t test for independent samples.
Significant at the .05 level for a tvo-tailed test.
** Significant at the .01 level for a tveotailed test.












TABLE 5


STRONG "?" SCORES OBTAINED FROM THREE GROUPS OF SUBJECTS,
iAtL AND FEMALE, SELECTED ON THE BASIS OF NMPI
ITEM OMISSION BEHAVIOR


HCS KCS LCS
GROUP GROUP GROUP
N M SD N N SD N M SD

Males 64 10.33 4k43 69 8.k2 4.Oz 56 8.54 3.69

Females k3 7.47 2*.8 0k 7.18 2.85 38 5.11 30o


TABLE 6

ANALYSIS OF VARIANCE OF MALE "?" SCORES OBTAINED FROM THE
STRONG FOR THREE GROUPS OF SUBJECTS SELECTED ON THE
BASIS OF MMPI ITEM OMISSION BEHAVIOR



Source of Variation SS df Variance Estimate

Between Groups 146 2 73.00

within Groups 3.139 186 16.88

total 3.285 188


F 4.32 (Significant at the *05 level)












TABLE 7


MEAN DIFFERENCES BETWEEN STRONG "?" SCORES FOR THREE
GROUPS OF MALE SUBJECTS SELECTED ON THE BASIS OF
MMPI ITEM OMISSION BEHAVIOR



Groups MI M2 t a


HCS vs LCS 1.79 2.32 *

HCS vs MCS 1.91 2.59
LCS vs MCS .12 .17


a t test for independent samples.
* Significant at the *05 level for a two-tailed test.
t Significant at the *02 level for a two-tailed test.



TABLE 8

ANALYSIS OF VARIANCE OF FEMALE "?" SCORES OBTAINED FROM
THE STRONG FOR THREE GROUPS OF SUBJECTS SELECTED ON
TE BASIS OF MMPI ITEM OMISSION BEHAVIOR



ource of Variation SS dt Variance Estimate

between Groups 123.74 2 61.87

within Groups 913.82 109 8.38

otal 1,037.56 111
I n j i .


F 7.38 (Significant at the .01 level)









-47-


TABLE 9

MEAN DIFFERENCES BETWEEN STRONG "7" SCORES FOR THREE
GROUPS OF FEMALE SUBJECTS S'ECTED ON THE BASIS OF
MMPI ITEM OMISSION BEHAVIOR



Groups M1 M2 to

HCS vs LCS 2.36 3*92 ***

HCS vs MCS .29 .38

LCS vs MCS 2.07 2.99 **


a t test for independent samples.
e** Significant at the .001 level for a two-tailed test.
** Significant at the *01 level for a two-tailed test.



TABLE 10

ANALYSIS OF VARIANCE OF INTOLERANCE OF AMBIGUITY, IN
TERMS OF MALE AND FEMALE CET SCORES, FOR THREE
GROUPS OF SUBJECTS SELECTED ON THE BASIS OF
MMPI ITEM OMISSION BEHAVIOR



Source of Variation SS df Variance Estimate

Rove (Sex) 37*38 1 37*38

Columns
(No. MMPI Items Omitted) 11.67 2 5.84

Interaction 4.82 2 2.41

within Cells 1,422.73 84 16.94

total 1,476.60 89

Interaction Effects F .14 (Not significant)
Column Effects (No. MMPI Items Omitted) F *45 (Not
significant)
Row (Sex) Effects F 2.21 (Not significant)













TABLE 11


ANALYSIS OF VARIANCE OF DESIRE FOR CERTAINTY, IN
TERMS OF MALE AND FEMALE DC SCORES, FOR THREE
GROUPS OF SUBJECTS SELECTED ON THE BASIS
OF MMPI ITEM OMISSION BEHAVIOR



Source of Variation 5 df Variance Estimate

Rows (Sex) 1558,933.61 1 1,558,933.61

Columns
(No. MMPI Items
Omitted) 379,728.29 2 189,864.15

Interaction 1,597,995.54 2 798,997.77

Within Cells 33,94,213.86 84 398,123.97

Total 36,979,071.30


Interaction Effects F 2.01 (Not significant)
Column Effects (No. MMPI Items Omitted) F = *&8 (Not
significant)
Row (Sex) Effects F 3.92 (Not significant)



TABLE 12

TEST OF THE SIGNIFICANCE OF THE DIFFERENCE BETWEEN
THE MEAN NUMBER OF MMPI OMISSIONS FOR TWO GROUPS
OF MMPI ITEMS CATEGORIZED ON THE BASIS OF
SOCIAL DESIRABILITY RATINGS


N M SD t

avorable Items 221 9.10 9.24
2.06 *
unfavorable Items 345 7.59 7.99


a t for two independent samples.
Significant at the .05 level for a two-tailed test.














APPENDIX B


TEST INSTRUMENTS



Desire for Certainty Test


Instructions Indicate on the answer sheet provided you
the probability (some number front 0 to 100) you oeel is
appropriate to the following statements* Indicate also
by circling one of the numbers (1, 2, 3, and 5) the
degree of certainty you feel with regard to your
probability estimate. Do not spend too nuch time on any
one statement and finish as quickly as you can. Roemnber,
each probability estimate refers to so many chances out
of 100.

1. The chances that an adult American sale will earn at
least $4,000 a year are about**. in 100.

2. A student entering law school will quit before getting
his law degree o**.

3. Frequent thumbsucking during childhood will make the
teeth stick out (causing buck teeth) ***.

4~ The President of the United States will be a man with-
out a college education ..*0

5. A major league baseball team will win the pennant if
it is in first place July kth o**.

6. A sexual pervert will have a low intelligence (IQ 80
or less) )o**

7. A high school graduate will go on to a freshman year
in college oo*

8. A couple getting married this year will later have a
divorce o...

9. An American sale now at the age of 40 will live beyond
the age of 55 ****
-&9-









-50-

10. An American family will live in a place without a
telephone *seo

11. An American family will own its own home **..

12. The telephone number you call will be busy *o..

13. An American citizen will believe in God ....

l. A varsity football player in an American university
will be subsidized (given money for his football
ability) 0o..

15. An American city of over 50,000 people will have a
chapter of the League of Women Voters sees

16o The governor of a state will be elected for a
second term in office **o.

17. A son will go into the same kind of work as his
father o**.

18. A man 70 years old will need financial help from
someone to support himself **o*

19. Spanking a child will make him tell the truth the
next time *...

20. An American-born baby will get a poor and inadequate
diet during his first year of life **..

21. An adult male will stay home instead of going to
church on Sunday ....

22. A sixth grade teacher in the public schools will be
a San **o*

23. A child whose parents are divorced will be
neurotic ***.

24. In the United States a girl will be married by the
age of 17 .*o*

25. A world champion boxer comes from a poor family ....

26o An American citizen will be bilingual (speak two
languages) **eo

27* A five card deal will have two cards of the same
kind (one pair) ....

28. A man with a broken neck will die o**.








-51-

29. A crime in the United States will be solved (someone
arrested and convicted for it) ****

30. The number of aute accidents in a year will be higher
than for the year Just before ....

31. A small business (for examples gas station, notel)
vill fail within two years after starting ..0.

32. The person one marries will have the same religion ....















CONCEPT EVALUATION. TECHNIQUE


Card No.,
Concept No.,
and Card
Orientation

I. 1 A
A
2 A

3 A
A


IIe









Tile









IV*


Concept


bat

fish

witch


Location Subject's Response
(Beck) Yes No


W

DI

DZ


4 person D3

5 V man W



6 A wvoan Dd2Z

7 Atwo people W
8 A dancer Ds5

9 >dog D
10 A butterfly D3



11 A tree D0
12 A face D8
13 A butterfly D3

14 A two people Dl

15 A turkey D2


16 A person DI

17 A man W
18 A candle D3
19 A shoe D6
20 A animal skin W


I


I I I L I












Yes No


V. 21

22

23
242


A
A


VI. 25

26

27

2


A


VII.











VIX.









IX.


bat


person's leg
man


man

dog's face
man' ; face

animal skin


Dl


D3


Ds

V/


0 A child Dd22
30 river DC

31 A two women D2
32 V shrub 3

33 A frog



3 A ribs. D3

35 two people D0

3 A rainbow V

37 A animal DI



3, two witches D3

3 A alligators D5
0 A fro_ D1

I1 V va:e Ds3G

A2 boat DC

3 > man's head D:


*White space immediately above left D1.


-53-








5t4.


Yes No
I I | i i i i i l i


rabbit's head D5

intestines D11
crab or octopus D1

flashlight Dd2&

two women DA

skeleton Dds0O

pitcher DZ


Total Number of Rejections


X0 44

41;
)& 6

47

Jks

49

50


A

A

A

A

A


___ __ _ I __ __~


---






-- --- --














APPENDIX C


NUMBER AND PERCENTAGE OF UNIVERSITY OF FLORIDA FRESHMEN WHO
VERE HIGH CANNOT SAY SCORERS GIVING CANNOT
SAY RESPONSES TO EACH ITEM



Book- No. No. Total 5 Males % FoPales % Total
lot Males Females No. Oaitting Omitting Omitting
No. Omitting Omitting Omitting
(N 68) (N 35) (N 103)


-55-











Appendix C-continued


Book- No, No. Total % Males % Females % Total
let Males Females No. Omitting Omitting Omitting
No. Omitting Omitting Omitting
(N 68) (N 35) (N 103)


-56-








-57-


Appendix C--continued


Book- No. No. Total % Males % Females % Total
let Males Females No. Omitting Omitting Omitting
No. Ottin Ottig itng Omitting
(N 68) (N 35) (N 103)

77 6 o 6 9 0 6
78 10 2 12 15 6 12
79 7 3 10 10 9 10
80 1 1 2 1 3 2
81 6 6 12 9 17 12
82 5 2 7 7 6 7
83 0 1 1 0 3 1
84 7 2 9 10 6 9
85 1 0 1 1 0 1
86 10 4 1A 15 11 1f
87 1 2 3 1 6 3
88 3 3 6 t 9 6
89 15 6 21 22 17 21
90 0 0 0 0 0 o
91 5 0 5 7 0 5
92 2 2 4 3 6 4
93 12 7 19 18 20 18
94 7 3 1 10 9 10
95 3 0 3 4 0 3
96 2 3 5 3 9 5
97 4 2 6 6 6 6
98 24 10 34 36 29 33
99 7 3 10 10 9 10
100 6 5 11 9 1k 11
101 19 8 27 28 23 26
102 12 4 16 18 11 16
103 5 0 5 7 0 5
10o 0 1 1 0 3 1
105 4 O k 6 0 4
106 1 1 2 1 3 2
107 2 3 5 3 9 5
108 8 1 9 12 23 9
109 2 1 3 3 3 3
110 5 0 5 7 0 5
111 3 1 4 4 3 4
112 7 5 12 10 14 12
113 1 0 1 1 0 1
11k 0 0 0 0 0 0
115 18 6 24 27 17 23
116 3 5 8 4 1t 8
117 1k 2 16 21 6 16
118 0 0 0 0 0 0
119 5 1 6 7 3 6
120 0 0 0 0 0 0










Appendix C--continued


Book- No. No. Total % Males % Females % Total
let Males Females No. Omitting Omitting Omitting
No. Omitting Omitting Omitting
(N 68) (N = 35) (N 103)

121 0 0 0 0 0 0
122 2 0 2 3 0 2
123 0 0 0 0 0 0
124 13 7 20 19 20 19
125 0 0 0 0 0 0
126 10 1 11 15 3 11
127 20 7 27 30 20 26
128 5 0 5 7 0 5
129 11 1 12 16 3 12
130 0 1 1 0 3 1
131 t 3 7 6 9 7
132 6 1 7 9 3 7
133 1 k 5 1 11 5
134 6 3 9 9 9 9
135 7 2 9 10 6 9
136 6 1 7 9 3 7
137 1 3 a 1 9 4
138 3 1 4 4 3 4
139 3 I1 3 4
1o0 5 1 6 7 3 6
141 7 1 8 10 3 8
142 2 0 2 3 0 2
143 6 1 7 9 3 7
1kk 8 k 12 12 11 12
145 2 0 2 3 0 2
146 6 2 8 9 6 8
147 4 3 7 6 9 7
148 2 0 2 3 0 2
k9 0 0 0 0 0 0
150 3 3 6 i 9 6
151 0 0 0 o o 0
152 3 1 a i 3 4
153 0 0 0 0 0 0
15; 0 1 1 3 1
155 2 1 3 3 3 3
156 0 0 0 0 0 0
157 2 1 3 3 3 3
158 2 0 2 3 0 2
159 1 1 2 1 3 2
160 11 7 18 16 20 17
161 2 1 3 3 3 3
162 13 4 17 19 11 17
163 9 0 0 13 0 9
164 7 6 13 10 17 13
165 10 3 13 15 9 13








-59-


Appendix C--continued


Book- No. No. Total % Males % Females % Total
let Males Females No. Omitting Omitting Omitting
No. Omitting Omitting Omitting
(N 68) (N 35) (N 103)

166 4 o 4 6 0 4
167 7 1 8 10 3 8
168 5 1 6 7 3 6
169 0 0 0 0 0 0
170 5 0 5 7 0 5
171 5 2 7 7 6 7
172 9 2 11 13 16 11
173 7 1 8 10 3 8
174 1 1 1 0 1
175 0 0 0 0 0 0
176 1 0 1 1 0 1
177 9 3 12 13 9 12
178 0 0 0 0 0 0
179 1 2 6 6 6 6
180 3 0 3 4 0 3
181 7 1 8 10 3 8
182 3 1 & 3 4
183 13 2 15 19 6 15
184 0 1 1 0 3 1
185 1 0 1 1 0 1
186 1 0 1 1 0 1
187 1 0 1 1 0 1
188 4 1 5 6 3 5
189 1 0 1 1 0 1
190 0 1 1 0 3 1
191 3 0 3 t 0
192 1 0 1 1 0 1
193 0 0 0 0 0
194 5 0 5 7 0 5
195 2 1 3 3 3 3
196 3 o 3 4 3
197 2 0 2 3 0 2
198 4 1 5 6 3 5
199 10 5 15 15 14 15
200 3 0 3 4 o 3
201 11 6 17 16 17 17
202 3 0 3 4 0 3
203 6 2 8 9 6 8
20z 3 1 4 & 3 4
205 3 1 4 4 3 &
206 6 0 6 9 a 6
207 1 0 1 1 0 1
208 8 1 12 12 11 12
209 9 5 11 13 14 1i
210 0 0 0 0 0 0







-60-


Appendix C--continued


Book- No. No. Total % Males % Females % Total
let Males Females No. Omitting Omitting Omitting
No* Omitting Omitting Omitting
(N 68) (N 35) (N 103)

211 0 1 1 0 3 1
212 2 0 2 3 0 2
213 2 1 3 3 3 3
214 2 0 2 3 0 2
215 0 0 0 0 0 0
216 7 1 8 10 3 8
217 1 2 3 1 6 3
218 1 1 1 0 0 1
219 0 4 0 11 k
220 8 4 12 12 11 12
221 3 0 3 i 0 3
222 12 6 18 18 17 17
223 10 8 18 15 23 17
224 1 1 2 1 3 2
225 4 o 4 6 0 4
226 5 0 5 7 0 5
227 0 0 0 0 0 0
228 9 2 11 13 6 11
229 6 3 9 9 9 9
230 2 1 3 3 3 3
231 26 5 31 39 1J 30
232 17 7 24 25 20 23
233 14 8 22 21 23 21
234 4 1 5 6 3 5
235 7 1 8 10 3 8
236 10 2 12 15 6 12
237 34 16 50 51 46 49
238 4 1 5 6 3 5
239 10 5 15 15 1 15
2o0 2 1 3 3 3 3
2l41 8 1 9 12 3 9
242 2 6 3 11 6
243 2 1 3 3 3 3
24z 13 3 16 19 9 16
245 2 2 3 6 h
246 3 0 4 0 3
247 1 0 1 1 0 1
248 6 6 12 9 17 12
2k9 17 8 25 25 23 24
250 10 6 16 15 17 15
251 2 0 2 3 0 2
252 7 1 8 10 3 8
253 10 1 11 15 3 11
25z 12 4 16 18 11 16
255 15 10 25 22 29 24










Appendix C--continued


Book- No. No. Total % Males % Females % Total
let Males Females No Omitting Omitting Omitting
No. Omitting Omitting Omitting
(N 68) (N = 35) (N 103)

256 1 0 1 1 0 1
257 6 5 11 9 14 11
258 10 2 12 15 6 12
259 5 1 6 7 3 6
260 1 5 6 3 5
261 6 1 7 9 3 7
262 5 2 7 7 6 7
263 2 0 2 3 0 2
264 6 3 9 9 9 9
265 3 2 5 4 6 5
266 10 3 13 15 9 13
267 5 2 7 7 6 7
268 16 3 19 2& 9 18
269 4 0 4 6 0 4
270 5 2 7 7 6 7
271 9 3 1 13 9 12
272 3 1 h 1 3 4
273 3 0 3 A 0 3
274 2 1 3 3 3 3
275 3 0 3 4 0 3
276 -k 2 6 6 6 6
277 4 2 6 6 6 6
278 1 2 6 6 6 6
279 1 0 1 1 0 1
280 10 4 1b 15 11 14
281 7 0 7 0 a0 7
282 2 3 5 3 7 5
283 6 1 7 9 3 7
284 11 5 16 16 14 16
285 2 2 4 3 6 k
286 7 1 8 10 3 8
287 33 13 A16 9 36 45
288 1 1 2 1 3 2
289 10 8 18 15 23 17
290 5 2 7 7 6 7
291 0 0 0 0 0 0
292 1 0 1 1 0 1
293 1 0 1 1 0 1
294 0 0 0 0 0 0
295 15 5 20 22 14 19
296 10 0 10 15 0 10
297 19 12 31 28 34 30
298 14 7 21 21 20 20
299 26 12 38 39 34 37
300 10 0 10 15 0 10








-62-


Appendix C--continued


Book- No. No. Total % Males % Females % Total
let Males Females No. Omitting Omitting Omitting
No. Omitting Omitting Oaitting
(N 68) (N 35) (N 103)

301 0 1 1 0 3 1
302 0 1 1 0 3 1
303 3 1 4 I 3 4
o3A 3 0 3 o0 3
305 1 1 2 1 3 2
306 9 3 12 13 9 12
307 0 2 2 0 6 2
308 5 1 6 7 3 6
309 3 2 5 4 6 5
310 6 8 14 9 23 Ih
311 3 0 3 0 3
312 2 0 2 3 0 2
313 6 2 8 9 6 8
314 7 3 10 10 9 10
315 1 0 1 1 0 1
316 11 3 14 16 9 14
317 10 14 15 11 14
318 5 2 7 7 6 7
319 17 1 18 25 3 17
320 8 0 8 12 0 8
321 5 1 6 7 3 6
322 4 1 5 6 3 5
323 5 2 7 7 6 7
32g 2 3 5 3 9 5
325 2 0 2 3 0 2
326 1 0 1 1 0 0
327 6 2 8 9 6 8
328 4 0 4 6 0 a
329 3 1 3 4
330 o 1 1 o 3 1
331 7 3 10 10 9 10
332 2 1 3 3 3 3
333 3 1 a 4 3
334 4 1 5 6 3 5
335 2 2 h 3 6 k
336 2 1 3 3 3 3
337 6 3 9 9 9 9
338 7 1 8 10 3 8
339 2 0 2 3 0 2
340 2 0 2 3 0 2
341 0 2 2 0 6 2
342 4 0 6 0 4
343 4 3 7 6 9 7
34h 2 1 3 3 3 3
345 0 0 0 0 0 0










Appendix C--continued


Book- No. No. Total % Males % FoMales % Total
let Males Females No. Omitting Omitting Omitting
No. Omitting Omitting Oattig hitting
(N 68) (N 35) (N = 103)

346 2 1 3 3 3 3
347 10 2 12 15 6 12
348 8 4 12 12 11 12
349 6 0 6 9 o 6
350 2 0 2 3 0 2
351 2 1 3 3 3 3
352 3 &4 b 3 4
353 3 2 5 & 6 5
35 o0 2 2 0 6 2
355 1 1 2 1 3 2
356 5 0 5 7 0 5
357 5 3 8 7 9 8
358 9 1 10 13 3 10
359 3 5 8 & 14 4
360 1 0 1 1 0 1
361 7 0 7 10 0 7
362 10 2 12 15 6 12
363 1 1 2 1 3 2
364 9 1 10 13 3 10
365 0 1 1 0 3 1
366 4 3 7 6 9 7
367 2 3 5 3 9 5
368 9 2 11 13 6 11
369 9 2 11 13 6 11
370 3 1 4 3 &
371 6 2 8 9 6 8
372 11 6 17 16 17 17
373 18 6 24 27 17 23
374 3 2 5 4 6 5
375 7 1 8 10 3 8
376 12 3 15 18 9 15
377 7 3 10 10 9 10
373 6 2 8 9 6 8
379 7 1 8 10 3 8
380 6 3 9 9 9 9
381 5 1 6 7 3 6
382 17 10 27 25 29 26
383 5 1 6 7 3 6
386 3 3 6 1 9 6
385 1 1 2 1 3 2
336 6 2 8 9 6 8
387 20 11 31 30 31 30
388 0 1 1 0 3 1
389 4 6 10 6 17 10
390 14 10 24 21 29 23







-64.


Appendix C--continued


Book- No. No. Total % Males % Females % Total
let Males Females No. Omitting Omitting Omitting
No. Omitting Omitting Omitting
(N 68) (N 35) (N 103)

391 10 4 14 15 11 14
392 1 0 1 1 0 1
393 5 0 5 7 0 5
394 7 1 8 10 3 8
395 4 3 7 6 9 7
396 4 2 6 6 6 6
397 3 1 4 4 3 1
398 4 0 4 6 0 4
399 1 0 1 1 0 1
a00 30 13 43 45 36 42
401l 1 1 0 3 1
o02 a 1 3 3 3 3
103 8 3 11 12 9 11
Lho 17 7 24 25 20 23
o05 0 0 0 0 0 0
406 7 3 10 10 9 10
407 3 1 4 4 3 4
408 6 3 9 9 9 9
4a09 1 5 6 3 5
410 16 5 20 2b 14 19
4llI 2 2 1 3 6 &
412 1 1 2 1 3 2
413 22 9 31 33 26 30
414 5 4 9 7 11 9
415 26 18 44 39 50 13
416 4 3 7 6 9 7
417 5 2 7 7 6 7
418 1 3 4 1 9 k
419 0 0 0 0 0 0
o20 5 2 7 7 6 7
421 1 1 2 1 3 2
422 2 1 3 3 3 3
423 2 2 h 3 6 4
424 1 1 2 1 3 2
425 3 1 1 4 3 &
426 8 1 9 12 3 9
427 4 6 10 6 17 10
428 11 3 14 16 9 14
429 14 1 18 21 11 17
30o 6 2 8 9 6 8
431 4 5 9 6 14 9
432 3 1 4 4 3 &
433 2 1 3 3 3 3
43) 6 3 9 9 9 9
435 21 7 28 11 20 27







-65-


Appendix C--continued


Book- No. No. Total % Males % Females % Total
let Males Females No. Omitting Omitting Omitting
No. Omitting Omitting Omitting
(N 68) N 35) (N 103)

436 13 11 24 19 31 23
437 18 7 25 27 o20 2
438 7 7 1b 10 o2 14
439 3 1 4 4 3 4
440 7 1 11 10 11 11
4h1 l 13 27 21 37 26
kB2 1 1 2 1 3 2
l43 10 7 17 15 20 17
Ah 11 7 18 17 20 17
445 10 5 15 15 14 15
446 2 & 6 3 11 6
447 8 6 14 12 17 14
448 2 0 2 3 0 2
449 5 1 6 7 3 6
a50 11 1 12 17 3 12
451 9 4 13 13 11 13
452 3 2 5 a 6 5
A53 6 1 7 9 3 7
454 9 0 9 13 0 9
B55 15 5 20 22 1t 19
456 1b 8 22 21 23 21
457 4 0 4 6 0 4
h58 3 1 a 4 3 4
59 0 1 1 0 3 1
460 2 0 2 3 0 2
b61 5 9 7 11 9
b62 1 0 1 2 0 1
463 13 1 1b 19 3 14
S64 9 t 13 13 11 13
465 1o 0 10 15 0 10
466 0 0 0 0 0 0
467 0 1 1 0 3 1
h68 12 3 15 18 9 15
h69 8 5 13 12 14 13
470 2 f 6 3 11 6
471 10 b 14 15 11 14
472 b 3 7 6 9 7
473 6 1 7 9 3 7
47h 6 0 6 9 0 6
475 21 12 33 31 34 32
476 9 4 13 13 11 13
477 21 16 36 30 46 35
478 8 a 12 12 11 12
479 3 0 3 b 0 3
480 2 0 z 3 0 2







-66-


Appendix C--continued


Book- N No No. Total % Males % Females % Total
let Males Females No. Omitting Omitting Omitting
No. Omitting Omitting Omitting
(N 68) (N 35) (N 103)

481 2 1 3 3 3 3
482 2 0 2 3 0 2
483 26 7 33 39 20 32
484 12 9 21 18 26 20
485 17 12 29 25 34 28
486 1 1 2 1 3 2
487 3 3 6 4 9 6
488 1 0 1 1 0 1
489 7 5 12 10 14 12
A90 2 0 2 3 0 2
491 14 5 19 21 14 18
492 5 2 7 7 6 7
493 17 9 26 25 26 25
494 1 1 2 1 3 2
495 18 11 29 27 31 28
496 6 9 15 11 14 11
497 4 3 7 6 9 7
498 3 2 5 4 6 5
499 1 3 4 1 9 4
500 5 5 10 7 11 10
501 6 4 10 9 11 10
502 6 4 10 9 11 10
503 8 6 14 12 17 14
504 17 7 214 25 20 23
505 3 1 4 1 3 4
506 11 3 14 18 9 15
507 8 8 16 12 23 16
508 1 0 1 1 0 1
509 8 4 12 12 11 12
510 3 3 6 4 9 6
511 & 2 6 6 6 6
512 1 0 1 1 0 1
513 30 13 43 45 37 42
511 6 5 11 9 14 11
515 0 1 1 0 3 1
516 4 1 5 6 3 5
517 1 0 1 1 0 1
518 6 & 19 9 11 10
519 2 1 3 3 3 3
520 8 4 12 12 11 12
521 6 2 8 9 6 8
522 0 0 0 0 0 0
523 9 0 9 14 0 9
521 5 1 6 7 3 6
525 1 2 3 1 6 3
526 1 1 2 1 3 2







-67-


Appendix C--continued


Book- No. No. Total % Males % Females % Total
let Males Females No* Omitting Omitting Omitting
No. Omitting Omitting Omitting
(N 68) (N 35) (N 103)
527 3 1 3 3 3 3
528 7 2 9 11 6 9
529 5 2 7 7 6 7
530 1 0 1 1 0 1
531 9 3 12 1I 9 12
532 10 5 15 15 14 15
533 3 0 3 a 0 3
534 16 8 2t th 23 25
535 1 0 1 1 0 1
536 3 2 5 6 5
537 2 0 2 3 0 2
538 4 0 I 6 o0
539 1 2 3 1 6 3
5o0 0 0 0 0 0 0
5I1 2 1 3 3 3 3
542 5 3 8 7 9 8
5.3 0 0 0 0 0 0
54s 1 1 2 1 3 2
545 2 0 2 3 0 2
506 1 3 h 2 9 h
547 5 h 9 7 11 9
548 6 9 15 9 6 15
549 & 5 9 6 1 9
550 11 3 1i 16 9 13
551 6 2 8 9 6 8
552 2 2 k 3 6 h
553 1 2 3 1 6 3
554 12 3 15 18 9 15
555 0 2 2 0 6 2
556 3 1 4 4 3 4
557 2 0 2 3 0 2
558 22 10 32 33 29 31
559 1 1 2 1 3 2
560 2 0 2 3 0 2
561 10 1 11 15 3 11
562 23 9 32 sk 36 31
563 12 4 16 18 11 16
564 5 i 9 7 11 9
565 2 1 3 3 3 3
566 12 k 16 18 11 16














BIOGRAPHICAL SKETCH


Morris Lee Raddy was born October 20, 1935, at

Bushnell, Florida. Following graduation from Bushnell High

School in June, 1953, he enrolled in the summer session at

the University of Florida# He received the Bachelor of

Science degree from the University of Florida in June, 1957.

He enrolled in the Graduate School of the University of

Florida in September, 1957. and received the degree of

Masters of Rehabilitation Counseling in January, 1959.

In February, 1959, he began work toward the degree of

Doctor of Philosophy in the field of cliniael psychology

and received this degree in February, 196o2

While in the Graduate School he worked as a graduate

assistant at the University of Florida Reading Laboratory

and Clinic and as a graduate assistant in the Department

of Psychology* For one semester he was employed part time

as a clinician in the University Counseling Center* Ho

served a one year internship in clinical psychology at the

Columbus Psychiatric Institute and Hospital in Columbus,

Ohio from February, 1961, to February, 1968.

Morris Leo Saddy is married to the former Noras Loe

Blanks of Miami, Florida. He is a member of Phi Gamma

Delta fraternity and is a student Journal member of the

American Psychological Association.

.68-

















This diamertation vL: preQ'ilOC under the direction of

the chairman of the candidate#* supervisory committee and

has bees approved by all members of that omauittee. It

as submhitted to the Doen of the College of Arts and

goiences and to the Oraduote Counei1, and vas appreved as

partial fulftlilelet of the requirements foZ the degree of

Doctor of Philosophy*

Februsry 39 1962






Dean, ra ge ate Arts and -9oheonoe




Sees# Graduate 44'ageol --


Supervisory Committee$


chairman

-/ c




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs