l"|

l

Hlll

\
1
1

W

J

HUM?

\
‘

_{
I

O)_\

m-t

(1)010

 

PREDECTEQR GE MALADAPTEVE
RESE‘QN‘SES UNDER CGRDETEGNS G?
HAaéT-WWEFERERCE FiiOM
RORSCHACH COLE}? RESPGNSES

Thesis k‘zr H16 Degree of Pix, 1?.
MiCHiGIAN STATE? C-SLLEGE
Robafi‘ W. Haz'fingi‘en

”i953

 

 

.i
i .
\ ,.
e \
I
1,
I
x
/\
1
,
h ‘
r
( \
' _ t 1 l
a ‘\
7 fl
1
. ‘ ‘
4
l
‘ J
\ '. .
\
, .
t _
1 ..
a I ,
v
, i if s,
, 4,.
v
IA 1
i x ,
.
i
/
v
I
l .
» s
I g ,
V
.. ‘
r J ‘
‘ v
, n
A ’ /
\_
I .
v
4
I " ,
I .1," «
> ‘A
I‘ ' / ,
. .1 , M V

This is to certify that the

thesis entitled

presented bg

has been accepted towards fulfillment

of the requirements for

degree in

”We

[/ Major professor

 

 

PREDICTION OF MALADAPTIVE RESPONSES ULDLR CONDITIONS OF
HABIT-INTERFERLNCE FROM RORSCHACH COLOR
RESPONSES

BY

Robert W. Harrington

Submitted to the School of Graduate Studies of Michigan
State College of Agriculture and Applied Scienco
in partial fulfillment of the
requirements for
the degree
of

Doctor of Philosophy
Department of Psychology
1953

{a

/ s

'5th:

 

 

lrjfi/i'l/

t‘
(4

ii
ACKNOWLEDGEMENTS

The author is indebted to many persons for the help they so.
freely gave during the course of this investigation.

To the members of the Guidance Committee, particularly Dr.
Alfred Dietze and Dr. albert I. Rabin, I wish to express my ap-
preciation for their patience, encouragement and constructive
criticisms.

It would have been impossible to carry out the study without
the help of the staff of Boys Vocational School. Mr. Robert Wiener,
Superintendent, made special efforts to help; Mr. E. L. V. Shelley,
Chiengsychologist, graciously provided access to the clinic files
for needed data; ur. Harold Smith, Mr. Donald Parker, Mr. Leo Norris,
and Mr. Paul Spata also gave freely of their time and effort. To
all of these and to the other personnel who aided in various ways
the author wishes to express his gratitude.

v Last, but not least, a debt of appreciation is acknowledged

to the boys who served as subjects in any of the phases of the

study.

3.531887

PREDICTION OF TALiDAPTIVE RQSPO7353 Uflilfi CJTDITIOES
OF HABIT-INTERFEREWCE FROM RORSCHACH

CCLOR RESPONSES

By

Robert W. Harrington

AN ABSTRACT

Submitted to the School of Graduate Studies of Michigan
State College of Agriculture and Applied Science
in partial fulfillment of the
requirements for
the degree
of

DOCTOR OF PHILOSOPHY

Department of Psychology
Year 1953

Approved @fiw)” %A’Q§:

The need for adequate experimental validation of
Rorschach hypotheses has long been realized by cost
users of the test. Regardless of the apparent useful-
ness of these hypotheses in dealing with individuals,
attempts to test them in clinical use remain subject
to errors of observation, judgement and preconceived
ideas. While the application of experimental proced-
ures to such problems does not necessarily eliminate
these errors experience has shown that they nake such
errors less likely to occur.

One of the major propositions of the Rorschach
test is that the type of rcSponse given to color is
indicative of the level of emotional deveIOpment of
the subject. Specifically, it is held that the FC re-
sponses indicate a higher level of emotional develop-
ment, iao., emotional maturity, than do CF and C re-
sponseo.

This investigation attempted to put this hypo-
thesis to test by controlling a number of other vari-
ables and using response to color as the independent
variable and reaction to conditions of frustration as
the dependent variable. A major assumption underlying
the investigation was that there is a positive re-
lationship between emotional maturity and adequacy of

reaction to conditions of frustration. This assumption

was shown to have empirical as well as fornal validity.

The experimental design consisted of selecting two
groups of subjects who were equated on several Rorschach
variables as well as age and IQ. These groups were se-
lected so that they differed on the independent variable
'of type of color emphasis on the fiorschach test. One
group, called the FC group, had given color responses
which were primarily of the F0 or form-dominant type.
The other, designated the C/CF group, had given color
responses in which the emphasis was predominantly on the
color element, i.e., C and/or CF resnonses. This pro-
vided two groups who differed on the variable of emotion-
al maturity, as indicated by color responses on the Ror-
schach, while being equivalent in regard to other vari-
ables.

Each group was subjected to conditions of frustrat-
ion and measures taken of their reaponses to these condi-
tionI. Habit-interference served as the frustration
variable by interfering with the goal—directed ;ehavior
of the subjects. The prediction was that the emotional-
ly immature group, i.e., the C/CF group, would show the
greatest degree of impairment in p rforrance under con-
ditions of frustration. This prediction was clearly
borne out on one of the tasks used (code-substitution)

and six of the seven ueasures taken on the second task

V— -|' v.4"V

‘ .

o"

-1,

&
.Ydibfisv IGVfiE' vt Lisa 3? iFDi“wiu ~v a; ’a.:e

owe gnknobloe lo hefictaece nylesl lean twenxh edl

J 4
doeriéndfi is'mvsr. ":1 is: none mm- om am:

we W €47

h .L , l
-.§‘sfibt lice?” ssedfl .91 “fit '2 er [[h' a: F-'

sfdshtv Josi--ne3shn? s ‘ nc 15-15133, .wn Jr - was be:-

etfl .deeJ deansr'mn art: no elem. helm "r-
eesnouews Trier nevi: r.“ ,qnm1z 01 an: he‘l-o ’

ll .Ostd dnnfl‘ nb-nfio‘ 10 33 a ' 1a ytiu t1"
. solos (as-vi; hex! ,quo-xp 138.5 9‘: r-sutrvglnslv .1136...“

.413 no (lflfirrrf-‘nobeT‘ new Etc—1N .1 .".d , . wane—-

em sttfl' .neam.» ten": '11 save». 3.3 ,nn p.310 '1'.

'e

6.90330“ To Oldnt'rfiv o n“. 12‘5’70'11 L .-. sv‘un'r-l out? be

. z .5103 on: no suitcases ‘qu: v.1 to 11>." _‘ ,vsrwae
' “31916013 [“6351 at Jil‘ . m .a more/f 9m» ,ds.
.8-

431M! ‘5 saustbnc- o: bisoefidus sew :‘flb‘u do”
Abuse seen: a: «success 1!: I' '1. ashlar ve-waaav I‘nu ‘
stalwart: ed: as some: 2' penchant-aids“ .lm
. 10W With-51393 em 1: '1» who?» at 1-: side?
elm lid and! In noiaotbu‘m all! .elostdu so:
-sd'3 fl Mm ,4”; 135.3 adv "4.1 ,nvreszr. o-w‘mi
,;:W‘M1"q ~k anezfiinvt Jo sonsb swam
, V; Ma‘ububer um .aotaem ‘0 Idol;
N WMlm-ohsl beau lino-r m" to one no inc em

    
 

J In: housed: no ads: es'xueser never eds 'ko It I

3
were in the predicted direction. It is felt that the

Rorschach assumption relating the types of color re-
sponses to varying levels of emotional develOpment is
supported to some extent by these findings.

It was suggested that the t*pe of methodology used
in this etudy could be extended to deal with more com-
plex types of behavior such as are met with in the

clinical use of the Rorschach test.

 

 

 

 

Clan-4. A‘.-..’1 . n A

TABLE or couTluTs
Amuscdioshimrs
LIST CF TABLES
LIsT<Avriouu55
INTRCUUCTIUN

Laotiohal maturity
Color Rosyouses and Emotional maturity

Emotional naturity and Reactions to Conditions

of Frustration
Conditions of Frustration
habit-interference

HYPUTHLQIS Cb Thh sTUUY
MATHUD
SUBJfiCTS
The Code-substitution Test
The Sirror-truciug Test
TRCCLDChm °
The Code—substitution Test
The Mirror-tracing Test
Instructions
nNALYSId “ND hLSULTS
The Code-substitution Test
The Mirror-tracing Test
DISCUSSILH OF REJLLTS
SUMLARY “NU CUECLUJICNS
APPENDIX A
A General survey of Validity Studies on the
Rorschach Test
APPLNUIX B
Rorschach Color hes,onses: nssumptious and

Validation Studies

APPENDIX C

COpies of Two Forms of the Code-substitution test

iii

ii

vi

Oaflvw \n

\O

10

18
18

a
0

19

f‘

4

2A

APPENDIX D
Photograph of the Mirror-tracing Apparatus

REFERENCES

iv

59
60

LIST CF TABLES

Table

1.

Comparison of age and IQ Data for Groups
Jith Primary Color emphasis (C/CF) and
Secondary Color Emphasis (F0) on the
Rorschach Test

Comparison of Rorschach Scores for Groups
With Primary Color smpLasis (C/CF) and
Secondary Color Emphasis (PC) on the
Rorschach Test

Mean scores on Each Trial of Two Forms of

a Code-substitution Test for Groups with
Primary Color Emphasis (CfCF) and Secondary
Color emphasis (FC) on the Rorschach Test

Comparison of Mean Impairment Scores for
Each Series of Three Trials on a Code-
Substitution Task for Groups With Primary
Color Emphasis (C/CF) and Secondary Color
Emphasis (FC) on the Rorschach Test

Comparison of Total Impairment Scores on a
Code-substitution Test for Groups with
Primary Color Emphasis (C/CF) and Secondary
Color Emphasis (PC) on the Rorschach Test

Comparison of Number of Errors Made on Two

Forms of a Code-substitution Test for Groups

dith Primary Color Emphasis (C/CF) and

Secondary Color Emphasis (FC) on the Rorschach

Test

Comparison of Performances on a Mirror-
tracing Test for Groups With Primary Color

hhphasis (C/CF) and Secondary Color Emphasis

(PC) on the Rorschach Test

Inge

12

13

29

33

vi

LIST OF FIGURES

Figure Page

1. Comparison of Performances on Two Forms
of a Code—substitution Test for Groups
With Primary Color Emphasis (C/CF) and
Secondary Color Emphasis (FC) on the
Rorschach Test 31

INTRODUCTION

Since it was introduced in 1921 the Rorschach test has re-
ceived a considerable amount of what may be called "clinical"
validation; i.e., it seems to work in practice. On the other hand
it has remained resistant to the more rigorous experimental vali-
dation desired for scientific acceptance. Although leading
Rorschach workers such as Beck (11), Benton (13), and Hertz (29)
have emphasized the need for the latter approach, relatively
little progress has been made in this direction.1

The present experiment was undertaken in an effort to add to
the scientific knowledge about the test by applying traditional
experimental procedures to the problem of the validity of one of
the major Rorschach dimensions--the color responses.2

The meaning of color reaponses. Rorschach assumed that

 

color responses on the test had meaning in reference to the emot-
ional life of the subject and stated, "The symptom value of the
color interpretations lies in the field of affectivity" (56, p.193).
The majority of workers following Rorschach have continued to work
on the basis of this original assumption. A further assumption
made by Rorschach was that the various types of color responses
were meaningful in relation to the degree of emotional adaptation

1 A review of validity studies on the Rorschach
test is presented in Appendix A.

2 A more complete discussion of color responses on

the Rorschach, together with a review of perti-

nent validation studies, is presented in Appendix

B.

 

 

of the individual. Thus, the C and CF reaponses were held to be
indicative of the degree of emotional instability, impulsivity,
suggestibility, sensitivity, lability, and irritability of the
person producing them. The FC reaponses, on the other hand, were
said to be characteristic of the normal individual who is well—
adjusted and able to make new adaptations in emotional behavior.
This latter type of color response, the F0, was held to indicate
a combination of affective and intellectual adaptability.
Following the orientation provided by Rorschach, Beck (8)
arranged the types of color responses in an order which he con-
sidered to correspond to the stages of "emotional maturity" (i.e.,
the level of emotional development) of an individual. He conceiv-
ed the levels of emotional growth as ranging from the lowest, at
which there is little or no control over the feelings, to the
highest at which there are only intellectual responses. Beck
arranged the color reaponses in the corresponding order of C, CF,
FC, and F}. He considers the F/ response to be the principal in-
dex of the ability of a subject to "...direct his thinking from
his higher centers..." (8, p. 19). Beck considers it as reflecting
conscious control and respect for reality and goes on to state,
"It is a measure of the strength with which the personality
has organized its values, the force that holds it tOgether
when disruption threatens, and the central directive to which
it refers at moments of decision....The F/ score becomes thus
the measure of the stability of the personality" (8, p. 22).

A further discussion of F as an indicator of control is

presented in Appendix B.

 

          

3.; .

In regard to the meaning of the various color reaponses,
Hertz and Baker state,

"The CF and C together reflect egocentricity and the more
primitive, childlike emotions, hence immaturity. Thus, emo-
tional stability and adaptability vary directly with the exp
cess of FC over CF and C. The smaller the proportion of FC
in relation to CF and C the more egocentric and unadaptive
the emotionality" (31, p. 13).

an important aspect of the representative vieWpoints pre-
sented above is the idea that color reaponses provide information
about the level of emotional maturity of the person producing
them.

Emotiong; Maturity. The concept of emotional maturity is
poorly defined in the literature. Saul (60) presents a discussion
of the concept in which he lists eight characteristics of the
fully mature individual. Recognizing the fully mature person as
an ideal he lists the following as characteristics of this end-
point of develOpment: (a) independence and self-reliance, (b) the
ability to give rather than receive, (c) freedom from feelings of
inferiority, egotism, and competitiveness, (d) socialization,

(e) integrated sexual functioning, (f) creative rather than de-
structive expenditure of energy, (g) a firm sense of reality, and
(h) flexibility and adaptability.

The view which Saul presents is essentially that as one

achieves maturity he relinquishes earlier modes of behavior and

does not fall back upon them in times of stress. The emotionally

immature person, on the other hand, either fails to develop to

A
higher levels or this deveIOpment is incomplete so that under stress
the earlier forms of reaction reoccur. Many other writers on the
subject of personality development agree on certain types of be-
havior which characterize the emotionally mature in contrast to
the emotionally immature person (2, 33, 48, 68).

Emotional maturity is usually defined in terms of a de-
velOpmental scale ranging from infancy to adulthood-with certain
types of behavior being considered appropriate or characteristic
for a lower level but not for a higher level. Thus, most writers
who have dealt with the problem agree that as a person grows from
childhood to adulthood he Optimally acquires more control over
his emotional reactions. at the lower levels of development these
reactions are said to be relatively gross, undifferentiated, and
changeable. There is also agreement that with increasing maturity
these reactions become more stable and differentiated; likewise,
the reactions become more specific to and appropriate for the
situations in which they.are called forth. Thus, to refer to a
person as being emotionally mature means, simply, that he char-
acteristically behaves in certain ways such as those described
above. Since the mature individual does not remain, in Klopfer's
words, "a slave to his feelings," nor is he usually thrown vie--
lently off balance by them, he is able to engage in emotional be-
havior and relationships which are typically neither over-inhibited
nor grossly exaggerated.

In discussing emotional development most writers use such

5
terms as "emotional stability," ”Well-adjusted," "emotional bal-
ance," "emotionally adaptable," and "emotionally healthy" as being
either synonymous with or reflecting emotional maturity. In the
material to follow it is to be understood that the use of such
terms includes this element of emotional maturity or lack of it.

Color responses and emotional maturity. Rorschach workers
maintain that a corresponding develOpment occurs in the individ-
ual's performance on the Rorschach test. The position taken is
that as an individual matures emotionally this development is re-
flected on the Rorschach test in that he no longer reacts with the
pure C reaponse, representing uninhibited emotional expression,
but integrates an element of form into his percepts. The CF and
FC responses, in which progressively greater emphasis is placed
upon the form element, are said to reflect increasing degrees of
control over emotional reactions.

Emotional maturity and reactions to conditions 2; £523-
tration. Another common element in the various definitions of
emotional maturity concerns the manner in which persons react to
conditions of frustration. There seems to be general agreement
that as an individual passes from one level of emotional develop-
xnent to a higher one his ability to maintain the efficiency of his
adjustment under conditions of frustration willshow improvement.
TTNAs, the person at a low level of emotional development will
‘38nd to react with more random, maladaptive, types of behavior in

tJie face of frustrating conditions than will a person who has

6

attained a higher level of deve10pment. As Symonds says, regard-
ing reSponse to frustration in relation to personality development,

"another criterion of the adequacy of reactions to frus-

tration can be stated in terms of the progression or regress-
ion of these reactions....Where learning takes place we see
evidences of growth and greater adaptability. This is a sign
of increasing maturity" (68, p. 67).

The assumption that emotionally immature persons will show
more-maladaptive responses under conditions of frustration than
will the more mature plays a cardinal role in this experiment and
hence deserves further discussion. On the one hand the validity
of this assumption is a purely formal matter since emotional ma-
turity is defined partly in terms of adequacy of reaction to condi-
tions of frustration. On the other hand this assumption is sup-
ported by empirical evidence.

Pertinent to the problem are studies such as that of
Zander (78) who demonstrated that emotionally unstable persons
reacted aggressively to a smaller degree of frustration and inter-
ference than did emotionally stable persons. additional support
for the assumption is furnished by Sherman and Jost (63) who com-
pared the reactions of normal children and adults and a group of
emotionally disturbed children to frustrating conditions in a
Ilearning situation. Measures of physiolOgical responses indicated
‘that the neurotic individuals reacted more intensely to the frus-
txrating conditions than did the normal. Similar findings are re-
INDrted by Malmo, et.al. (45) and Miller (47). Moreover, Zander

(78) has summarized the points of agreement in frustration studies

and concluded that the ability to tolerate frustration (i.e.,
maintain efficiency of behavior under conditions of frustration)
is an indicator of a healthy personality.

In view of this general agreement regarding the relation-
ship between emotional maturity and reaction to conditions of
frustration this relationship provides us with the basic tool for
the present investigation.

One result of accepting the above premise is that it sug-
gests the possibility of testing the Rorschach assumption relating
the several types of color response to verging levels of emotional
development. This assumption may be experimentally tested in the
following way:

Two groups of subjects, who, on the basis of their
Rorschach color responses, differ in their degree of emo~
stional maturity are subjected to conditions of frustration.

By definition, and on the basis of empirical evidence, the
emotionally immature persons will be empected to show the
greatest degree of disruption in behavior under such condi-
tions. This breakdown in behavior, therefore, provides a
basis upon which to compare the groups. The prediction is
that the group which gives "immature," i.e., C and/or CF color
reaponses will show greater disruption of behavior than will
the group giving more "mature," i.e., PC, color reaponses.
In this way disruption of behavior under conditions of frus-
tration presents itself as a manageable variable for testing
the Rorschach assumption under question.

Conditions of frustration. In this study "conditions of
:frustration" or "frustrating conditions" refers to interference
V'ith goal-directed behavior. This interference with goal-directed
behavior constitutes a condition of frustration which can be ex-

IDerimentally manipulated. This usage of the term "frustration" is

ill accord with that of Barker (A), Kardiner (3h), McGeoch and

Irion (43), Masserman (L6), murphy (L9), and Symonds (67).

Although most common, this is not the only way the term
"frustration" has been used in the literature. It is, therefore,
important to keep in mind that it is the "interference with goal-
directed behavior," and not some inner state in response to such
conditions, which is herein referred to as "conditions of frus-
tration" or "frustrating conditions." As will appear later, our
research design utilizes habit-interference as the independent
frustration variable which interferes with the goal-directed be-
havior of the subject.

Habit-interference. This construct, "Habit-interference,"
has been used to eXplain the typical negative effect found when a
subject, after learning a response to a stimulus, is required to
learn a new response to the same stimulus. .This negative effect
has been found with various types of learning tasks including
code-substitution, verbal learning, and motor tasks. After build-
ing up a strong habit in response to a stimulus the subject is re-
quired to learn a new habit in response to the same stimulus in
this t;pe of study. The old reSponses seem to persist and to in-
terfere with the accuisition of the new habit. It is apparent,
then,_that habit-interference may be considered to establish a
(Handition of frustration since it constitutes a condition in
Wftich tiere is interference with the goal-directed behavior of a

subject attempting to learn a new response.

HYPOTHEsIs or THE STUDY
After the prececding review and discussion of the concepts
employed in this experiment we may state the hypothesis of the
study as follows:

Individuals whose color responses on the Rorschach test
are predominantly of the C and/or CF type will show greater
impairment in performance under conditions of frustration,
i.e., interference with goal-directed behavior, than will
those whose color reSponses are primarily of the FC type.

This hypothesis is derived from the Rorschach assumption

relating the types of color response to varying levels of emotion-*
al maturity and from the assumption that there is a positive re-
lationship between emotional maturity and adequacy of reaction to

frustrating conditions.

10

METHOD

The first step in testing this hypothesis consisted of se-
lecting two groups of twenty subjects each. These groups, as will
be seen later, were equated on a number of variables considered to
be relevant to the problem. The groups were selected so as to dif-
fer on the independent variable of the type of color reSponses
given on the standard Rorschach test. One group, designated as
the FC group, was composed of subjects whose color responses were
primarily of the FC type. The other, called the C/CF group had
given color reaponses which were predominantly of the C and/or CF
type. This provided two groups who differed on the variable of
emotional maturity, as measured by Rorschach color responses,
while being eruivalent on other relevant variables.

Both groups were tested on two eXperimental tasks in which
habit-interference served to establish conditions of frustration.
The response to these frustrating conditions, then, served as the
dependent variable upon which the two groups were compared. The
prediction was that the "emotionally mature" group, i.e., those
whose color responses were primarily of the FC type, would perform
‘more efficiently under conditions of frustration than the "emotion—
ally immature" group, i.e., those who gave predominantly C and/or

CF types of color responses.

ll

SUBJECTS

The sample consisted of forty, white, male, juvenile delin-
quents. These subjects had been committed to Boys Vocational
School, Lansing, Michigan, a cottage-type institution with a popu-
lation of approximately 350 to LOO boys.

TWO groups of twenty subjects each were equated on the
factors of age and IQ as well as several Rorschach variables.
Table 1 presents the data on age and IQ. In all cases the age re-
ported is the age to the nearest half-year at the time of testing
on the experimental tasks. It can be seen from Table 1 that with
means of 15.85 and 15.75 years, respectively, the FC and C/CF
groups did not differ in age. The estimate of intelligence is re-
ported in the form of IQ's obtained from individually administered
Nechsler¥Be11vue Intelligence scales, Form I or II. This test had
been given shortly after each boy arrived at the institution.

The range in IQ was from 73 to 128 for the two groups comp
bined with a mean for the FC group of 102.75 and for the C/CF
group 101.35. Thus, it appears from the data in Table I that the
two groups did not differ significantly in intelligence as meas-
ured by the scales used.

In addition to age and IQ, the two groups were equated on
Several Rorschach variables, namely: F/Z, number of M, total Y,
ntuuber of P, and total number of responses. Table 2 presents the
dilta on the Rorschach variables. The independent Rorschach vari-

atile was the ratio: Number of FC reSponsss/Number of C/CF responses.

12

TABLE 1
Comparison of Age and IQ Data for Groups With Primary
Color Emphasis (C/CF) and Secondary Color Emphasis (F0)
on the Rorschach Test

(N 20 in each group)

 

 

Group
FC 0,10?
Mean Range Mean Range
Age 15.85 13.0-17.0 15.75 13.5-18.0
IQ 102.75 76-128 101.35 73-125

 

Ap.op eodaauaozpv Hoo. use» .aoa no m a nonsense“ **
Asa.» eoaaxpuozuv do. can» mmoa we a a nepxoaeea *

l3

 

**m0.a oo.~ as.a oo.a om.H oo.m on a.nesz
**Ha.a ma.m mN.H em.m mm. mm. as g.ne:z
*mN.m mo.H ms.a mo.a oo. oo. o u.naaz
HH.H we.“ me.HH ma.ma mm.m os.am axe
«a. no. mq.m mm.m ao.m om.m Adapeev
m u.nasz
>0. mm. mm.a oa.H NH.H ma.a =_a.neez
om.H om. ew.H om.e am.~ ow.e a gonasz.
mo. «0. ma.oa mm.m~ mo.m oo.m~ e aaaos
a .aaae am a an a
homo

nsouu

 

 

Adsoum new. ca om 2v

pqoa counonuom esp co momv qumnqsm aoaoo
humusooom use AmQXQV naaandam guano mumaflhm
new; gauche you nohoom noanoanom we confiquSoo

N mqmda

1h
As a general rule this ratio had to be at least 1/2 for inclusion
in the FC group and 2/1 for the C/CF group. Table 2 also presents
the data for the groups on the color responses. The color respons-
es are reported in unweighted form. It can be seen that the two
groups were roughly equated in regard to the frequency of all the
factors considered with the exception of the color variables.
The average ratio of FC/C/CF was 3.0/.35 for the FC group and
1.0/3.55 for the C/CF group. It should be pointed out that the
separate types of color reSponse were not assigned weighted scores
in arriving at this ratio. The ratio was arrived at by adding the
number of each type of color response given by the group and divi-
ding by the number of subjects in the group. For individual sub-
jects the range on this ratio was from 2/0 to 0/0 for the FC
group and from 0/2 to 5/8 for the C/CF group. No subject had a
ratio which was different in direction from that of his group.

All Rorschach reaponses were scored by the investigator
and the scoring of the color reaponses was checked by another
psychologist. Non-color responses about which there was some
scoring question were similarly checked.

Equating groups of this size on the above-mentioned Ror-
schach variables, in addition to age and IQ, was a difficult task.
There were two factors in the situation, however, which made the
task less formidable than it first appeared. Each boy had pre-
viously been given a group Rorschach test as well as an individ-

ual intelligence test as part of the intake procedure of the

15
Psychological Clinic at the school. These records were available
to the investigator. The first step, then, consisted of screen—
ing approximately L00 group Rorschach records and selecting those
which contained color reSponses. The next step was to screen this
group of potential subjects for age, IQ, and Race. Following this
individual Rorschach examinations were given and the final groups
selected. To obtain the final two groups approximately 120 indi-
vidual Rorschachs were given. It is felt that without the screen-
ing procedure outlined above this number would have been much high-

er.

16

.IATERIALS AND APPARATUS

In addition to the group Rorschach and Intelligence tests,
which had been previously given, three tests were used. In the
selection of subjects the standard Rorschach test was individual-
ly administered and scored according to Beck's norms (7). Two
tasks were used to compare the perfornance of the groups under con-
ditions of frustration. One consisted of two forms of a code-
substitution test1 and the other of an electrically wired, six~
pointed star design in a mirror-tracing apparatus.

The Code-substitution Task. This consisted of two forms,
a and B. Each form contained a code consisting of geometric
figures with nunbers within each figure. Below the code, on
each form, appeared a number of blank figures which were to
be filled in according to the code given. The figures were:
star, circle, snuare, cross, and triangle. The same figures
appeared on each form. The numbers to be used to fill in the
figures were: 1, 2, 3, A, 5. Thus, the same figures and the
same numbers were used on each form but the number to be
placed in each figure differed on the two forms. Copies of
these forms are presented in appendix C.

The task of the subject was to put the appropriate number
in each figure as rapidly as possible. The instructions were
to start at the upper left-hand corner and do each figure in
turn until reaching the lower right-hand corner. Perfornance
was scored for time per trial and number of wrong substitut-
ions .

 

The Mirror-tracing 2325 involved tracing a sinpointed star
design under conditions of mirror-vision. The apparatus is
shown in Appendix.D. The design was constructed from copper
plates and a chronograph, counter, and stylus were connected
to these plates so that when contact was made between the
stylus and the edges of the pathway, i.e., when the subject

1 The forms are patterned after those included in
The Record Blank for The Revised anmination for
the Measurement of Efficiency of mental Function-
ing by Babcock—levy. These record blanks can so
obtained from C. R. Stoelting Co., Chicago, Ill.

17

left the pathway, the circuit was completed. This arrange-
ment provided a measure of the total time Spent off the path-
way as well as the number of times the subject left the path-
way. In addition, a stop-watch was used to measure the total
time spent in tracing the design.

l8
PROCEDURE

The Code-substitution Test. On the code-substitution test

 

each subject was given 18 trials on Form A followed by 12 trials
on Form B. Eighteen trials on Form A were considered sufficient
to develOp a strong habit and 12 trials on Form B sufficient to
reveal the differences in performance which were expected. The
addition of more trials on either form was contra—indicated by
two considerations: (a) it was necessary to keep the task from
being too long so that the interest of the subject would be main—
tained, and (b) on the basis of prior experimentation 18 trials
on Form a and 12 trials on Form B proved to be sufficient.

When the subjects were required to learn the second form
they had to learn to make new responses to old stimuli. This, it
was assumed would yield negative transfer or habit-interference.
Habit interference, by definition, is a condition of frustration
since it interferes with'the attainment of the goal of learning
the new reSponses. Since both groups could be shown to have
learned the first habit ecually well, and since motiVation was
assumed to be objectively equal for both groups, the frustrating
conditions could be considered the same for both groups. There-
fore, any difference in decrement of performance between the
groups on Form B could be attributed to greater inability of one
group to deal with the frustrating conditions in an adaptive man-
ner.

The Mirror-tracing Task. Following the administration of

 

19
the code-substitution task each subject was recuired to make one
complete tracing of the star-design pathway under conditions of
mirror—vision. The strength of the reSponse tendencies involved
in tracing a star-design under ordinary direct vision was assumed
to be the same for all subjects since there was no reason to sup-
pose differential anounts of CXperience in making eye—hand movements
in Space. Since these previously acquired response tendencies, or
habits, were assumed to be of ecual strength for the groups no
trlining was given to bui d up such a habit. This also provided
a situation which differed from that in the code-substitution task
since the two habits differed on the variable of reccncy of ac-
quisition. The rationale for this task was that under mirror-
tracing conditions, just as on the code-substitution task, the
previously learned reSponses could be expected to interfere with
the attainment of the goal and in this way provide another set of
frustrating conditions.

Instructions. Each subject was tested individually in a

 

room provided in the psychological clinic of Boys Vocational
School. The examiner met each subject, either in the waiting
room or in a cottage, and conducted him to the testing room. The
examiner had previously been employed at the school and had also
had prior acquaintance with the subjects in the cottages so that
after a few brief remarks, e.g., "Do you remember me?," "It's a
nice day," etc., It was possible to begin by saying:

"I suppose you are wondering what this is all about aren't

20

you? Well, this is something that has nothing to do with BVS.
I am trying to find out how well people can do certain things
and that's why I called you up here. This is my own personal
business and has nothing to do with BVS. I used to work here
a few years ago and because of that Mr. hisner (the super-
intendent) gave me permission to test Some of you boys.

I would like to have your help but you do not have to do
this if you don't want to. Nothing will happen to you if you
would rather not do it. In fact, unless you are willing to
try, and to do your best, I would rather you didn't do it be-
cause it won't help me unless you do your best. It will take
about an hour. Do you think you would like to?"

If the subject ansaered in the affirmative (no subject re-
fused or gave any other indication of not wanting to partici-
pate) E continued, "Fine, when we are all through if I think
you have done your best, and really tried, I want to give you
a candy bar for helping me out." (E pointed to the candy bars
which were in plain sight).

Placing a copy of Form A of the code-substitution test in
front of the subject E sai , "You may have had a test some-
thing like this before but this is a different one. Notice
the figures at the top. Notice that each has a different
number inside it. Here, below, are a lot of the same figures
and your job is to see how fast you can put the right number
in each figure. You start here (upper left) and end here
(lower right). You may not skip any figures and you must fill
in every one.

In case you make a mistake I will say 'wrong.’ If'you make
a mistake, do not stop to erase because that will only use up
time. If you make a mistake just write the correct number
over the wrong one you have put in the figure (E demonstrated).
If you make a mistAke and I do not see it, but you notice it,
correct it in the same Way.

When I say 'ready' you are to put your pencil-point inside
the first figure. when I say 'go' you are to begin filling in
the figures as rapidly as you can without making mistakes. I
will time how long it takes you from beginning to end. I will
not tell you how long you took each time but when you are all
finished you may see your scores if you want to. In case
your pencil breaks take another one and I will make allowance
for that time. No matter what else happens, if someone comes
in, or the telephone rings, keep right on working as fast as
you can.

We will do this three times and then rest, three times and
rest, three tLues and rest, and so on. He will keep doing it
until I think you have done it as well as you will be able to.
Do you have any questions? Remember, if you do your best you
get 8 Candy bar. Ready---go."

Each subject was then given 18 trials, in groups of three,

21
‘with two minutes betWeen each Set of three trials. During the two
[minutes between sets E engaged the subject in trivial conversation.
supproximately 10 seconds elapsed between the time the subject fini-
shed one trial, E recorded the time and identified the test sheet,
and S was presented with a new sheet and instructed to begin. Be-
fore each group of three trials B said, "Let's see how fast you
can do these."

Following the rest period after the 15th trial, B said,

"Now let's do the last three of this group and then do something
different." The subject was then given the last three trials.
after 1% minutes of the rest period following trial number 18 on
Form a had elapsed, B placed a copy of form B in front of the
subject and said,

"This is different in two ways. First, we will not do as
many of these as we did of the others, and secondly, it's
different in that the numbers have been changed around, see?
(8 pointed to the code). Otherwise it is just like before;
if you make a mistake, and I see it, I'll say 'wrong.’ If I
don't see it, but you do, correct it anyway. Remember to
work as fast as you can. Ready---go."

Except for the number of trials the procedure for Form B

was exactly the same as in the preceeding 18 trials on Form A.
Following 12 trials on Form B the subject was taken across the
room to the table on which the mirror-tracing apparatus was stand-
ing. Each subject was shown the apparatus and B said,

"Look at this star (direct vision). Your job will be to
start here (upper right-hand point) and trace the pathway as
fast as you can without touching the sides. You start here,

follow the arrows, and finish at the same place you started
from (E demonstrated). Every time you touch the sides this

22
counter clicks and this clock starts (B demonstrated). You
cannot get a shock from this even if you put your hands right
on the plate, see? (B demonstrated). If you touch the sides
come back into the pathway and keep going. Do not just slide
the stylus along the side but if you touch the sides come
back to the center and keep going.

Now you take the stylus and place it at the start (direct
vision). when I say 'go' you are to follow the arrow from
the start around to the finish as fast as you can without
touching tie sides. after I say 'go' the only place you can
look is in the mirror (design was then shielded from direct
vision). No matter what happens you cannot look anyplace but
in the mirror. Ready-~—go."

Many subjects reached a point where they apparently could
rust continue tracing and if they gave any indication of discontin-
tuing E said, "Just keep at it, you'll get on to it."

Following completion of this trsk E thanked each subject
arui gave him a candy bar as the reward which had been mentioned
iri the instructions. Each subject was requested not to discuss
tile tests with any of the other boys and was excused from the
room.

In both of the tasks it was assumed, as it is in most ex-
}Jezriments, that the instructions, in combination with the reward,
thald motivate the subjects. In this expe iment the fact that
bcrbh groups showed improvement in performance on Form A of the

Ccude-substitution task tends to support the assumption that the

Subjects were motivated.

23
ANALYSIS AND RESULTS

For the code—substitution task four comparisons were made
between the groups: (a) performance on Form A, (b) performance on
Form B, (c) impairment scores (B-A), and ((1) number of wrong; sub-
stitutions on each form.

In comparing the performance curves for the :groups simple,
non-parametric statistics were used. Fisher's "t" test (one-
tailed) was used to evaluate the differences netween groups in
impairment scores and number of errors made. The formulas employ-
ed were those given by McNemar (1.1., pp. 79, 226). In addition,
the F ratio was utilized to test for her-.mogeneity of variance and
Where the two groups were found to be significantly different on
this factor the correction indicated by Edwards (21;, p. 170) was
made in applying the "t" test.

The mirror-tracing test yielded three scores: (a) the num-
ber of times t ;e stylus left the pathway, (b) the total time Spent
off the pathway, and (c) the time required to trace the design.
The means, medians and stnx'xdnrd deviations of the groups were com-
puted for each measure and differences which appeared likely to
be significant were evaluated by means of the "t" test.

The results indicate that the code-substitution test re—
liably differentiated the two groups in the predicted direction
as measured by time scores. When compared on the number of errors
the trend was for the C/CF group to make fewer wrong substitutions.

At- first glance this finding would appear to be in the Opposite

21.
direction than predicted but since the task was to complete each
trial as fast as possible it does not necessarily follow that a
greater number of errors indicates poorer performance. This find-
ing is discussed more fully in the following pages.

While the mirror-tracing task revealed differences which
were mainly in the predicted direction these differences were not
statistically significant.

The Code-substitution task. 0n the code-substitution task
the two groups did not differ significantly on Form A since, as
the data in Table 3 indicate, the FC group performed better on 10
trials and the C/CF group on 8 trials. Table 3 also indicates,
however, that on form B the EC group obtained better scores on
.all 12 trials. If there had been no difference in performance
'between the groups they would have been expected to react on the
basis of chance in regard to whether one group or the other was
superior on any of the 12 trials. The FC group, however, made
superior scores on each of the 12 trials, the probability of this
occurring on a chance basis being .0002. With this small a prob-
ability it is reasonable to conclude that the two groups were not
reacting on a chance basis but actually differed in performance
on Form B. This result is taken to support the hypothesis that
the C/CF group would perform less efficiently than the FC group
under’the experimental condition of interference with goal-
directed behavior (i.e., conditions of frustration).

The breakdown in performance on Form B was analyzed by

TABLE 3
1
Mean Scores on Each Trial of TWO Forms of a Code-substitution
Test for Groups With Primary Color Emphasis (C/CF) and
Secondary Color Emphasis (PC) on the Rorschach Test

(N 20 in each group)

 

 

Group
PC czar
Form Form
Trial A B A B
1. 80.25 96.65 80.50 102.15
2. 70.a5 79.35 66.85 86.55
3. 65.35 77.70, 61.25 83.55
A. 58.65 75.30 56.30 78.25
5. 59.10 75.10 56.65 80.95
.6. 58.10 72.60 56.00 77.55
7. 52.90 69.20 50.35 72.45
8. 53.95 68.05 56.60 73.60
9. 52.80 66.35 52.50 73.60
10. h8.90 60.15 h8.75 69.75
11. 51.10 61.50 52.25 67.60
12. 51.40 60.15 50.80 66.20
13- £8.85 49.70
_16. 51.35 50.10
15. 51.00 51.10
16. 66.90 68.00
17- 2.8.1.0 1.8.1.5
i§;.____, 49.10 68.75

 

Scores in seconds

 

26
computing impairment scores for each subject. This score was ob-
tained by adding the scores for each series of three trials on
Form A and subtracting this sum from that obtained for each of the
correSponding series of three trials on Form B. The differences
between the groups in the magnitude of these impairment scores
were then tested for significance by means of the "t" and F tests.
A total impairment score, based upon all 12 trials, was similarly
computed and treated.

It can be seen from Table A that the mean scores of the
groups were significantly different, in the predicted direction,
on series I, II, and IV while the difference on series III ap-
proached significance. Table A also reveals that the groups dif-
fered significantly, again in the direction predicted, in variabil-
ity on series II, III, and IV. These results indicate that on
each of the four series the groups differed significantly, in the
expected direction, either in mean scores, variability, or both.
Table 5 presents the means and standard deviations of the groups
on total impairment score. The difference between groups on this
score is also statistically significant. These results lend fur-
ther support to the hypothesis that the CfCF group would Show a
Significantly greater impairment of performance under conditions
of frustration.

When the two groups are compared on the number of errors,
-i.e., placing a wrong number in a figure, it can be seen from

(“able 6 that there were no significant differences. If anything,

27'

moo. so do. no asau>_m concedes“ ****
No. no 05Ha> m nouaowbcd ***
mo. no osaa> m nonaoauca «t.
gave a uoaaautoao #

 

4<.wa 04.0m
am 5

ANH.HH.QHV>H

mo.mm mm.m4
am a

Ao.m.aVHHH

so.m~ mm.>4
an E

mo.m.avHH

A-Hafiauv noflaom

****oa.m ***ma.m *sxma.m mm.m
***®O.N Fm.H **om.H ****mN.m
mm.HN mN.mH oa.om On.om
mm.a4 mh.am mm.m¢ om.mb mo.a¢ m4.>o mo.mm mm.mo

54.NN m4.mm
am 3

Am.m.avH

*u

.HMHU

om

macho

 

 

Ansoum some cw om 2v

onus noncomuox ago so Aomv nfi-azasm uoaoo auuncooom use Amo\ov
cananasm uoaoo mumsaum 39H; nasoao you xnae :oHuSpfipunsuneooo u :0
nadwne «case we uoauom noafl new nouoom pecauamgsH sue: me sooauaasoo

¢ mqmde

TABLE 5
1

Comparison of Total Impairment Scores on a Code-
substitution Test for Groups With Primary Color
Emphasis (C/CF) and Secondary Color Emphasis (F0)

on the Rorschach Test

(N 20 in each group)

 

 

Mean ‘ sD diff. t
Group
FC 157.15 6a.12
. 90.20 2.h6*
c¢cr 2A7.35 1A6.76

5.24**

 

i indicates P of(.02).01 (one-tailed test?
** indicates P of .002

1

Scores in seconds

29

TABLE 6

Comparison of Number of Errors Made on Two Forms of a
Code-substitution Test for Groups With Primary
Color Emphasis (CKCF) and Secondary Color
Emphasis (PC) on the Rorschach Test

(N 20 in each group)

 

 

Form A Form B
M SD M SD
Group
FC 22.68 lh.63 31.26 19.38
c/CF 20.37 13.33 26.26 12.95
diff. 2.31 5.00

t 0 [4957 o 9107

 

30
the trend was for the C/CF group to make fewer of these errors.
This finding is interesting, in conjunction with the fact that
the groups differed in impairment scores, since it indicates that
while making the same or fewer number of incorrect substitutions
the C{CF group either had more difficulty in coping with these er-
rors, as indicated by their longer ime scores, or made more
errors of a different, more covert type. On an observational level
this type of error seemed to consist of vacillating movements to-
wards and away from the blank figures. A similar finding is re-
ported by Siipola (65) who found that in a code-learning problem
partial reSponses in the direction of the first learned response
made up more than two-thirds of all errors. Thus, the finding
that the C/CF group tended to make fewer wrong substitutions may
or may not be eXplained on the basis that they either took longer
to correct their wrong substitutions or made more errors of a dif-
ferent type. Only further experimentation can answer these ques-
tions .

It is also interesting, in regard to the hypothesis of the
C/CF group performing less efficiently on Form B, to compare the
shapes of the performance curves for the two groups shown in Fig-
ure 1. It is immediately apparent that the curves for the two
groups do not differ in shape but only in amplitude, 1.8., the
STOQPS did not differ in the rate of learning on this form but in

the<degree of learning as measured by time scores.

Seconds

110

100

80

7O

8

5O

40

3O

20

10

31

 

 

 

 

 

 

 

 

 

 

 

 

 

Test for Groups With Primary Color Emphasis (C/CF)

 

 

 

 

 

 

 

 

 

 

and Secondary Color Emphasis (PC) on
the Rorschach Test

 

 

 

 

 

 

 

 

T p . T.
| '. .u . 1
—H-[~~~-— c+ - ~> 4i. L 1 rrrrrr 4H . -, g ,- ‘ -- - p ...Ts... -—;—< o—.- »+‘ el
. ~v~ FTGURE l H .
”L -syLL La. ‘ H La... LLL. L_' i - ‘ L.“ Lutlgfiii L
La Comparison 0f Performances.on Two Forms 01 a Code-substitution

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. —— -————-—--_-— .— - qr..- .——o— -<— —- a... .— -—— Far- q————— -
.. . ‘, l.
b—o—oc k‘>-v——o—oo—<H—c-- - FV- - - —ck<-< -——-~—4p—~—. - .4 -VLH4 .1 -.- .1 4-“ Hat-1 H—O—H
, ,. . .. .
‘P- my PC Group (N 20, . . a
w-..—-A-—4~» c F sip m ’0) - .,
.. ‘ .! 4r .. .- ..
”-L‘VH... w-.—4>——.-- _4 o —o—4 ————4o—oo—H—4¢—~—o—Oi{
, ‘ .l .l.. ,,.. .
. ., .... .l.
. . .
. .. . ...
‘ L
., ....
.... ...l ,.
.... ... u.
. .. .‘J. . ..
>—~——-— Lo—c o—lL—a -r >—-—< m... - 1 ... -_. .Wa o—o—o—o—A
. . ..l, ...v
. .. ...‘ , . ..Ac
. ...A ,..¢
ll .... ...
,. .... . .
.. ... .... ....
. . ..l. 1...
- —----4 _-—---4—. .— -_. a-lp-«o—o—e-Awqw
. . ...
L ... .
\. .H.
. . V ... .
, L4 vv~ -I- i
Y -..- ——-—-{> M‘——-.
.. . V.o .
a: .
... 1
‘nl4 4
. i
.I. .
-..———-—4-.—.
. ... .
.. .
.o- -
-t- .
.. ... .
.. --- -
it ‘c-q a
, l , .,,_ . .
L
. ‘... .v, .. .
.1. ... .. .
.«c. .. ..-
..J: .... . .
L
‘3'. ... ....
‘ .... ..v ‘.,,
A. o.‘ ...- 41..
7 H“
5Ȣ- o... ....
lét| .... ..
0‘00 DUL- .
..- .,,. .
en ... .
...1 .
.. ... -
. ._.. .
b . ...l
.r:A vv“ ‘
.. 1.1- .
... l . . .... .
... “ .... .
..t .. . .
... l ‘ 0 ~ I: -
‘—‘—w‘ ¢~44r~—- 4-4
..L . ....
I I I n. I (CI. 4
. . .... .
.. l y , ..., A
.-- . . . . . .... ‘-
..‘ “ , .
. y , , . . ... ..
Lo—o-‘mo ->—¢ -4-aa—o—o a‘.-—4y- o-o—o-4 >-- v- 0*4H+Or‘ o a -‘w. 4L. m- . a 4 +4 } . . - }—- .. o~r+—¢_4.H >—o ‘—4 e—4r-< 0- -
" , . y _ , .... .
I p.. . ,, l
‘ ‘ ‘.,. . O"
. V .
0‘ . I . . . .... .. 1
l .,. ‘_ ‘V ' .. _, ‘
‘ 4 .. ... ,. . .llt .- .
. . |.A , -L . .. .. .
r-r—a—Ow ~< H... . *4 .5 ion -. .wb—O-o—o - .— H. L 4 oar . - A- . .“+<>—a buo—Q—rhtwc e . o .4 Q ~ L. e .. 7-, s4—.+ 0~u 0-0- .a o -o-4
-" .. ,n - V . .. .l .
.- , . .,. , . .. .. .,.. -- .
, ..V y A ‘-~- v' D
I .. .y .
.. a I-a- ..A.
I n: . ‘9'! (00"
.. .... , l. . .;.. ...l
. . , .... ... .. . ., ... ; .otv rv‘!
>— .-.4 .H.-. ..1.—.‘).._.- "9—. +4). Ho . -—-<.-‘..+ ”.4 e~ --... ‘H—< ..... A ~—. ........ —- u-o—e-orla—orw-od
' . c... ,‘,.,.... vn-a-l't
‘ o .- .. . . ...4 V.'
. ...- ...
L g ~.¢..-_-JL.~¢

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Trials

1 2 3 h 5 6 7 8 9 10 11 12 13 1h 15 l6 17 18

32

The Mirror-tracing Task. The means, medians, and standard
deviations of the groups1 on the mirror-tracing task are present-
ed in Table 7. While none of the differences between means or
{medians approaches statistical significance the C/CF group ob—
tained poorer mean scores on two of the three measures used and
poorer median scores on all of the measures. When the separate
scores were transformed into standard scores and an overall score
computed the C/CF group again obtained a poorer score which was,
however, not significantly poorer than that of the FC group. On
the whole, then, these differences represent trends in the pre-
dicted direction but it must be concluded that the two groups
could not be reliably differentiated on the basis of their per-
formances on the mirror—tracing task. A rank-order correlation
between performance on the code—substitution task and performance
on the mirror—tracing task fielded a Rho coefficient of -.06.

Evidently the performance on one task cannot be taken as a good
indication of performance on the other task.

Why performance on the mirror—tracing task failed to either
differentiate the groups at a statistically significant level or
to Show a significant correlation with performance on the code-
substitution task is an interesting question. Although the as-
sumption that there was interference with goal-directed buhavior

1 These figures are based upon an N of 19 in each
group. One subject in each group violated the
instructions and reverted to direct vision. Con-

sequently, these subjects could not be included
in the sample.

'33

oocosowmwo pcsowmwcmwm o: hau30fi>po on: open» ones; oofiamda no: out some a og9**
uedooom CH zo>dm we osfiat

 

an In ammm. Noo.H mmmm. In
oo.m mm.a om.oa sq.m 00.0 ow.
om.ma oo.dm om.4m 0H.04 05.0HH mm.waa 05.5m oo.Hm hm.00d
ou.aa oo.ma ma.mm om.mq Om.4m md.moa oq.mm oo.mm oa.aoa
mm .co: 2 am .Cnfi 2 an .com 2

*MMO oaae *oEHB deuce HMO uoEHH .02

**v
.mmao
moxo
om

nacho

 

 

Amzoum some cw ma 2v
puma sowsomnom esp co

Aomv nfiomzmam uoaou mpmocooom use AmQXUV ufimmzmam soaoo agsewbm
no“: museum new pace mcfiomauasounwz a co woodegomsom no comauagsoo

h m4m<a

' 31+
on both tasks seems defensible this does not necessarily imply
that the two situations were alike in other ways. It is, for ex—
ample, possible that different psychological processes were in-
volved in the two tasks. It is also possible that the interference
in the mirror-tracing task was easier to overcome. nnother possi-
ble explanation might be that the habits involved differed on the
variables of strength and/or recency of acquisition. A more plaus-
ible explanation, however, may be that not enough measures were
taken of performance on the mirror—tracing task. It is possible
that had more trials been given the effect of the frustrating con-
ditions on performance would have continued to be disruptive for
the CfCF group while the FC group would have overcome the disrupt-
ive effects more cuickly. Regardless of which explanation is con-

sidered, however, only further investigation can supply the answer.

35
DISCUSSION OF RfishLTd

« within the limitations imposed by the nature of the sample
and the tyre of activities measured, the results on the code-
substitution test are interpreted as lending some support to the
hypothesis of the study. Consequently, the Rorschach assumption
from which the hypothesis was derived is supported to some degree
by this experimental data. The results on the mirror—tracing task,
while in the predicted direction, do not support the hypothesis of
the study at a statistically significant level.

In this study it proved to be possible to predict, on the
basis of Rorschach behavior and in accordance with the assump-
tions of the test, how persons would react to conditions of frus-
tration. On the code-substitution task this prediction could be
made with a high degree of reliability while on the mirror-tracing
task the reliability of prediction was low. To the extent that
these results can be generalized to include reactions to a variety
of frustrating conditions the results increase the confidence
With which the clinician can employ the Rorschach test.

The fact that some of the results tend to support the vali-
<iity of a basic Rorschach assumption is not considered to be the
<3nly finding of the study. Methodologically the study would seem
tx: have implications for further work on the test in relation to
its use in clinical situations.

In clinical settings the aehavior of the subject, about

Wfiich the Rorschach worker is called upon to make predictions, is

. 36
very complex. There is seldom, if ever, any interest in how a
person will perform on a code-substitution test but predictions as
to how a person will respond to psychotherapy, what he will do
when angry, .nd similar predictions are commonly desired. These
predictions involve extremely complex relationships between what
the person does on the Rorschach test and his behavior in other

situations. Usually, however, in psychology as in other branches

u to the prediction

_)

of science, the discovery of the lans relatin
of complex events is preceded by the discovery of simpler relation-
ships and the evolving of methods applicable to the study of the
more complex events.

consequently, it is felt that one of the most fruitful re-
sults of the present investigation lies in the fact that it has
again been demonstrated, apparently, that isolated Rorschach de-
terminants can be dealt with in a carefully controlled, eXperi-
mental manner. A review of the literature reveals a limited, but
increasing, number of such attem ts. Taken together they hold
forth the promise of being able to circumvent the traditional
"global argument" of the Rorschach worker and to allow careful
study of the assumptions regarding the specific Rorschach deter-
minants. It is hoped that the present investigation may contri-
bute in some way to the solution of this broad problem.

The ”global argument" maintains that the meaning of any
determinant cannot be arrived at in isolation but must be inter-

Preted within the total complex of the Rorschach performance, e.g.,

37
the amount of M in relation to C is supposed to qualify the mean-
ing of C. Few experienced Rorschach workers would dispute the fact
that in any individual record the relationships between the various
scores are important. what is objected to by some psychologists
is not the "global argument" itself but that it has been used to
explain away negative results and as an excuse for not subjecting
tie deteruinants to rigorous experimental test. hore precise de-
finition of the meaning of the determinants should, in fact, make
the "global approach" to interpretation more effective since more
would be known about the variables which are being interrelated.
It would seem, then, that it is tine for the determinants to be
explicitly defined, in re; rd to their meaning, and subjected to
careful validation rather than to continue the above-mentioned
practices. It is lOSlCnl to assume that an increase in Knowledge
about the individual deter inants will facilit.te the discovery
of the interaction laws between them. since the "global approach"
is, essentially, an application of interaction laws, either ex-
plicitly or implicitly stated, such discoveries should improve the
value of this approach.

The present experiment seems to have demonstrated that the
individual determinants of the Rorschach are amenable to experi-
mental manipulation. It is hoped that further work will lead to
the develOpnent of other methods and techniques, as well as the
refinement of the present one, which may prove to be applicable to

the study of the more complex behavior patterns met with in

38
clinical situations. Ingram's study involving the use of white
space on the Rorschach (32) appears to mark a beginning in this
direction. In this study she measured the response to frustrat-
ion in an interpersonal situation by having several judges observe
and rate the responses of the subject when frustrated by another
person. The development of suitable techniques will probably not
be easy and will probably require the COOperation of workers in
other disciplines than gsychology. as Cronbach (22) emphasizes,
there is a need .or more appropriate statistical methods for deal-
ing with the Specific problems of the norschach and for this the
aid of the statistician will be needed.

Despite these dilficulties it would seem, however, that with
the development of adequate technigues of measurement and compu-
tation and the extension of the type of methodo105y used in the
present study, it flu; be possible to subject the individual de-
terminants of the test to Careful study and ultimately to dis—
cover their interaction laws. To the extent that these object-
ives are attained, the inte pretation of the test will rest upon
a sounder foundation.

In addition to these implications the results seem to be
. pertinent to the theoretical formulations of dchachtel (61).
Schachtel has postulated a relationship between the perception
of color and the expirience of affect in which a common element
is the "passivity" of the exp:riencin5 individual. By "passivity”

he refers to the degree of control which the subject has over his

39
reactions. as an illustration he compares the experiences under-
gone in a fit of rage and in the production of a pure C response
on the Rorschach. Schachtel concludes that in both there is a
relative lach of conscious control, direction, deliberation, and
detachment.

In the present experiment the group thich perceived the
relatively undifferentiated color on the horsehach exhibited less
control and direction of their behavior on the eXperLaental tasks.
Would these "relatively pure-color perceivers” also exhioit what
achachtel calls "passivity" in a variety of situations? If this
can be shown to be true under a variety of conditions and for di-
verse activities then the knowledge of the relationship between
perceptual and other nodes of behavior will have been advanced to
some degree. This may prove to be a fruitful problem for research
since the Rorschach test is founded upon the assumption that the
aetivity involved lies primarily in the field of perception. Con-
sequently, sound advances in perceptual theory should have import-

ant implications for the theoretical and practical aspects of the

AO
SUMMARY AND CONCLUSIONS

The need for adequate experimental validation of Rorschach
hypotheses has long been rralized by most users of the test. Re-
gardless of the apparent usefulness of these hypotheses in deal—
ing with individuals attempts to test them in clinical use remain
subject to errors of observation, judgement and preconceived ideas.
While the application of experimental procedures to such problems
does not necessarily eliminate these errors experience has shown
that they make such errors less likely to occur.

Cne of the major prOpositions of the Rorschach test is that
the type of reSponse given to color is indicative of the level of
emotional develOpmcnt of the subject. Specifically, it is held
that the FC responses indicate a higher level of emotional deve10p-
ment, i.e., emotional maturity, than do CF and C reSponses.

This investigation attempted to put this hypothesis to test
by controlling a number of other variables and using response to
color as the independent variable and reaction to conditions of
frustration as the dependent variable. a major assumption under-
lying the investigation was that there is a positive relationship
between emotional maturity and adequacy of reaction to conditions
of frustration. This assumption was shown to have empirical as
well as formal validity.

The experimental design consisted of selecting two groups
of subjects who were equated on several Rorschach variables as

Ewell as age and IQ. These groups were selected so that they dif-

Al

I

fered on the independent variable of type of color emphasis on

the Rorschach test. One group, called the FC group, had given
color reSponses which were primarily of the F6 or form-dominant
type. The other, designated the C/CF group3had given color re-
sponses in which the emphasis was predominantly on the color ele-
ment, i.e., C and/or CF responses. This provided two groups who
differed on the variable of emotional maturity, as indicated by
color responses on the norschach, while beins eouivalent in regard
to other variables.

Each group was subjected to conditions of frustration and
xneasures taken of th ir responses to "hese conditions. Habit-
interference served as the frustration variable b9 interfering
with the seal-directed behavior of the subjects. The prediction
was that the emotionally immature group, i.e., the C/CF group,
would show the greatest degree of impairment in performanCe under
conditions of frustration. This prediction was clearly borne out
on one of the tasks used (code-substitution) and six of the seven
measures taken on the second task (mirror—tracing) were in the
Predicted direction. It is felt that the Rorschach assumption
relating the types of color responses to varying levels of emotion-
al clevelopmtnt is supported to some extent by these findings.

It was suggested that the type of methodology used in this
sttuiy could be extended to deal with more complex types of behav-

lOF‘ such as are met with in the clinical use of the Rorschach test.

h2
AFFENDIX A

A GENLRAL SURVEY OF VnLIJITY STUDIES Uh THE RORSCHACH TEST

A great many validity studies have been reported in the
literature and no attempt will be made here to deal pith all of
these. Instead, this section is intended to provide the reader
with a general orientation as to the types of studies which have
been done. Representative examples are provided but for a more
comprehensive treatment of the problem of the validity of the
Rorschach the reader is referred to reviews by Benton (13), Hertz
(30), Rabin (53), and dilliams (76).

Types of studies and results. The methods that have been

 

most used in validation studies have been clas ified by Hertz (30)
as being of four main types: (a) direct experimentation with
changes in experimental conditions, (b) correlations of Rorschach
categories with objective behavioral criteria, (c) comparisons be-
tween groups, and (d) clinical case studies.

Rabin (53) points out that recent approaches towards vali-
dation have tended to be more strictly experimental and he dis-
tinguishes three varieties of such an approach. These three con-
sist of: (a) the molar approach in which the total personality
configuration is utilized, (b) the molecular approach which aims
tC) establish the validity of individual Rorschach factors, and
($3) the direct experimental approach which varies the experiment-
31 (Dr test conditiOns and/or measures physiological or organismic

COIKxxnitants within tie subject. Rabin emphasizes that these are

#3
not mutually exclusive categories but serve as classification de-
vices.

In dealing with the same problem, williams (7b) classifies
recent experimental approaches as: (a) comparison of Rorschach
reSponses with independent behavioral criteria, (b) correlation
of Rorschach response variation with variations of conditions dur-
ing administration of the test, and (c) the relating of Rorschach
responses to intervening or prior experiences which have been ex-
perimentally controlled.

among the more widely known validation studies are those of
Hilliams (75) on intellectual control under stress, Baker and
harris (3), who studied the Speech of college students under norm—
al and stress conditions in relation to test responses, and Brewer
(19), on cardio-vascular activity before and after visno-n tor
conflict. Other well-known studies are those of Bergman, Graham
and Leavit (15), th studied the effects of hypnotic age level re—
gression on nor chach resyonses, as well as the work of Lane (A0)
on the effect of hypnotiCally induced sets on the Rorschach move-
ment factor, and hilkins' and Adams' work with patients under hyp-
nosis and sodium amytal (7A).

In discussing eXperimental studies Rabin (53) lists several
under the molar approach including that of Benjamin and Ebaugh (12)
who compared the psychiatric diagnoses of patients with diagnoses
based on the Rorschach. They were able to demonstrate complete

agreement in about 85% of the cases. It is of note that in this

L4
-study the Rorschach diagnoses demonstrated greater agreement with
the final than with the preliminary psychiatric diagnoses. In a
similar study by diesel (6h), 26 children referred to a child
guidance clinic were given the Rorschach twice with an interval of
a year between tests. whereas initially there was only 61.5% ag-
reement with the psychiatric diagnoses after a years interval, in
which the psychiatric diagnoses were reevaluated, there was 86.5%
agreement. In another study of the molar type, Young and Higgen-
bothan (77) observed the behavior of unstable boys in a psychiat-
ric summer camp and were able to demonstrate some corresyondence
between Rorschach performance and such aSpects as intellectual
function, emotional factors, and interests.

as early as 1936 Thornton and Guilford (69) carried out a
study which has been characterized as_molccular. They atte pted
to correlate various Rorschach factors with scores on other per-
sonality measures. as Rabin (53) notes, confusion of definitions
and relitive inadeuuacy of criteria were weaknesses of this study
in which no significant relationships were found between the vari—
ablezs studied. Huesch and Finesinger (57), in another molecular
type of study, found that color scores, devised on the basis of
color type and size in drawings, correlated highly with the number
of color responses on the Rorschach. Thus, it was shown that the
individual's use of color was somewhat similar in the two situat-
ions.

The direct experimental approach has been used by Brosin

45
and Fromm (18), Kelley, Maréulies, and Barrera (36), Kelley and
Levine (35), Barbin (59), hilliams (75), and Rallen (73). Sarbin
(59) used hypnotically induced sets with a single subject and come
pared the Rorschach performances under these various conditions
with those obtained in a normal, waking state. Results indicated
that a number of Rorschach scores changed from one condition to
another but the meaning of these changes was not clear. Kelley,
margulies, and Barrera (36) compared the pre- and post-electric
shock Rorschach.performances of their subjects and concluded that,
on the whole, the aPElS and determinants were essentially stable.
The value of this method consisted of utilizing the amnesic state
following electric shock in order to compare the performances of
the same individual with the memory factor supposedly eliminated.

1 another experimental study carried out by Ingram (32) in-
volved the validation of the Rorschach white space as an indicator
of oppositional or agaressive tendencies. Taking two groups of
subjects, differing on the variable of high or low use of white
Space on the Rorschach, she placed them in situations of: (a)
standardised problem-solving, and (b) interviews. In both situa-
tions there was an attempt to frustrate the subjects and judges
rated their reactions. She found that the group which was high
on white Space was significantly more aggressive in the frustrat—
ing interview situation but that the use of white space did not
differentiate the groups in the problemesolving situation.

In an interesting experiment, Eichler (25) equated two

A6
groups on the Behn-Rorschach test and then, using the standard
Rorschach, tested one group in an anxiety-producing laboratory
situation and the other group under ordinary conditions. Of the
fifteen alleged Rorschach indicators of anxiety evaluated, he
found that only decrease in a, decreased productivity, increased
oligophrenic details, and increased shading differentiated the
groups.

Using post-hypnotic suggestion, Counts and Mensch (21) at-
tempted to influence the aorschach performance of five subjects.

They gave the sun estion of hostility and checked its effective-

(1
gt)

ness by observation and interview. Independently repeated Ror-
schachs, however, showed only trends in the expected direction.
Summary. In general it can be seen that, although vary-
ing in adequacy and in the nature of the results obtained, these
studies indicate some degree of validity for the test. Neverthe-
less, as Benton (13) notes, following a review of the literature
on the validity of five major aspects of the Rorschach, although
some progress has been made present day Rorschach interpretation
is based upon assumptions which have been inadequately validated.
One of these five aspects, the color responses, will be discussed

more conpletely in Appendix B.

147
APPENDIX B
RORSCHACH COLOR RESPONsRS: ASJUMPTIORS AND VALIDATION STUDIES

Rorschach's original scoring system classified color re-
sponses in one of three primary ways. A response was scored C if
the only deterainant was the color; a reaponse was scored CF if
the color was the primary determinant but form also played a part,
and a response was scored FC if the form was the principle de-
terminant but color was also important in the percept. Certain
assumptions were made concerning the psychological meaning of
these responses.

Rorschach's Assugptions. On the basis of his experiMents,
Rorschach concluded that the color reaponses had meaning in re-
lation to the affective life of the individual. as he expressed
it, "The symptom value of color interpretations lies in the field
of affectivity" (56, p. 193). In the use of the term "affectiv-
ity" Rorschach followed Blueler's usage and included the affect,
emotions, and feelings. The term "lability" as used by Rorschach
meant fluidity of the emotions and could include normal affective
adaptability. Also, in addition to the general significance of
COlor responses, Rorschach proposed meanings for the different
types of color reSponses.

Regarding the C responses, Itorschach thought they repre-
sented unstable emotionality and the more such reSponses given,
the more likely the individual would be to act impulsively:

"C answers are seen, first, in the irritable and sensitive,

A8

and increase in number in manics where irritability and in-
stability are features. Irritable and impulsive patients show
large numbers of C's as does the group of schizophrenics which
shows tte most emotional and associative scattering; epilep—
tics, also impulsive and irritable fall into this class" (56,
p. 31).

The CF reSponses were also held to be indicative of affect-

ive instability but were found to a considerable extent in normal

subjects,

for

"CF answers” wrote Rorschach, "prove to be analogous to
pure C answers, but occur in a larger number of subjects.
CF's are almost always found when C's are given but the in-
verse of this does not hold....They have proved to be the re-
presentatives of emotional instability, irritability, sensi-
tivity, and suggestibility" (56, p. 33).

For Rorschach, the FC reaponses represented the capacity
emotional adaptation and he wrote,

"The FC responses” are those interpretations which "re-
present the capacity for affective rapport and adaptability,
a kind of combination of affective and intellectual adaptabil-
ity. They are, in general, characteristic of the normal indi-
vidual who is well-adapted and is capable of making new adap-
tations" (56, p. 199).

Post-Rorschach assumptions: The Q Resnonses. Most Rorschach

workers have proceeded on the basis of the significance attributed

to these reaponsos by Rorschach. Hertz and Baker indicate,

"It may be concluded that if form represents conscious ex-
perience and reality, hence conscious control, the primary C
where no form is included, reveals lack of control over the
emotions, no feeling for the environment or reality, hence
inmmlsiveness" (31, p. 10).

Beck states, in discussing the C reSponse,

"It is an infantile response mode and thus not abnormal in
the very young child. Occurring in the older child and in the
adolescent, it leads to eXpectation of tantrums or other out-
bursts. The adult with pure C in his Rorschach record is like-
ly to oe given to ungovernable impulses..." (8, pp. 27-28)

49

For Piotrowski (52) these responses indicate emotional be-
havior which is impulsive and unsocialized, while Ueser (50) feels
that they indicate strong emotionality, distractability, and sug-
gestibility. Similarly, Schachtel and Hartoch (62) consider both
C and CF to indicate unstable, unbalanced emotions. Moreover,
Schachtel (61) considers that pure C indicates an ego which is
less able to exert control over internally originating affective
drives. The conclusion reached by Hertz and Baker, regarding C
reSponses, is that, "More than one indicates lack of control over
emotional reaction, lack of restraint, and impulsiveness” (31,

p. 57).

Thus, the psychological meanings attributed to the pure C
reSponse have included impulsivity, suggestibility, lack of rat-
ional control over emotional reactions, lack of social adjustment,
immaturity of affect, excitability, instability, sensitivity to
emotional excitation, and susceptibility to being easily upset.

in; 9: Response. according to Beck,

"The CF or color—form reSponse is characteristic for a

less impulsive but still highly labile reactivity. Deve10p-
mentally it marks a phase above that represented by C. ‘In it,
the infantile reaction trend is being held in restraint by
the F tendency, a force out of another psychic sphere—-aware-
ness of reality....The essential state is that of easy irri-
tability" (8, p. 28).

After reviewing several studies on the color responses,
Hertz and Baker (Bl) conclude that CF responses reflect states of
emotional instability and excitability in an individual and, like—

wise, KlOpfer and Kelley (38) maintain that these reSponses re-

 

 

 

 

50
flect emotional reactions which are more impulsive than those re-
presented by FC responses.

222 fig reshonse. RGéurdlfio these resforses KlOpfer and
Kelley (38) indicate that one of the best validated Rorschach as—
sunptions is that they are a measure of emotional adjustment to
outer reality. One characteristic of the emotionally healthy in-
dividual, according to Beck (8), is that he is able to control his
feelings. Beck considers the FC resyonse to be a messure of this
attribute. Hertz and Baker point out that, "FC represents emotion-
al stability and adaptability” (31, p. 13), and in the same way
Rapaport mainteins that the FC resnonses indicate "....flexible
control and careful regulation of the perceptual and associative
processes....” (5L, p. 236).

Post—Rorschach workers, it can be seen, have carried on
Rorschach's original assumptions regarding the meaning of the FC
response. This resyonse is held to reflect the emotional maturity,
the conscious control over emotions, capacity for affective rap-
port with the environment, adaptability and stability.

In addition to the individual color reaponses, Rorschach
emphasized the significance of the different amounts and combina-
tions of these in arriving at a description of a personality.

This emphasis has persisted in the efforts of later Rorschach work-
ers and can be seen in the statement of Hertz and Baker that,

"The CF and C together reflect egocentrici,y and the more

primitive, childlike emotions, hence immaturity. Thus, emot-
ional stability and adaptability vary directly with the excess

 

 

 

1 Ajw—q‘m.

51

“(fl

of FC over CF and C. The smaller the proportion of c in re-

lation to CF and C, the more egocentric and unadaptive the

motionality” (31, p. 13). '

Conceiving of the stages of emotional growth as ran;ing

from the lowest, at thish there is little or no control over the
feelings, to the other extreme at which there are only intellect-
ual responses, Beck (8} arranges the color responses in the cor-
responding order of C, CF, FC, and F/. These representative view-
points reflect the general assumption that where FC predominates
over C and CF, the individual is more emotionally adaptable, less
given to childish, innature behavior, is less unstable and in gen-
eral is able to deal with emotional stimulation in an effective
manner. Conversely, the "ssunption has been that where C and CF
predominate over PC, more childish, immature, unstable, and im-
pulsive behavior is expected and the individual is thought to be
less able to control the expression of his emotions.

Validation Studies. In aduition to the method of individ-

 

ual case study, attempts to demonstrate the validity of the as-
sumptions regarding the color responses have, in general, been of
four types: (a) group comparisons, (b) correlational studies,
(c) genetic and normative studies, and (d) experimental dttEthS.
Group Conparisons. Beck (5) compared two groups on the
basis of Sun C. One group was described as unstable and excitable,
and tge other as quiet, stable and adaptable. He reported a dif-
ference which, while not statistically significant, was suggestive

of the expected results. Hertz (28) found that an ascendent group

 

 

 

52
gave a greater Sum C than did a submissive group, and Loosli-
Ustori (42) found fewer FC and more CF in a poorly adjusted and
problem cases than in control groups of normals. These studies,
in general, support the Rorschach assunytions involved.

Comparing criminals, Boss (16) found that the more aggres-
sive types gave more C and CF and fewer FC reSponses. Somewhat
divergent results were obtained by doynton and walsworth (17) who
compared a group of reform-school girls with a group of high—
school girls and found that the high-school girls scored higher
in all respects in regard to color than did the delinquents. They
found that the impulsivity ratio (FC/CF/C) was more in favor of
C/CF in the high—school group. Pfister (51) obtained differences
between subnormal and clinical groups but was only able to find
agreenent with other clinical data in about fifty per—cent of the
cases. These latter two studies fail to furnish much support for
the underlying Rorschach assumptions.

Both Dubitscher (23) and Hertz (28) obtained a higher Sum
C with subnormal than with normal groups. Kerr (37) compared
normal, clinic, and mentally defective children and agreed that C
represents an impulsiveness which is usually pathological. Simi-
larly, Beck (6) and flickers (55) obtained more C and CF in their
schizophrenic groups than in control groups. Wallen (73) used
the group Rorschach with stable and unstable men in service and
found that the unstable group was more affected by color than

was the stable group. The results of these studies involving

 

53
group>conparisons have been generally in agreement with the Ror-
schaclxassunptions regarding color althoudh there have been some
exceptions (17, 51).

Correlational approaches. Hertz (28, 29) reported a cor-
relation of -.360 f .051 between the Hoodworth—Mathews Psycho-
Aneurotic Inventory and color on the Rorschach. Vernon (72) found
a correlation of {.47 between Rorschach scores and the ascendence-
submission test of allport and also reported a correlation of /.33
between Sum C and a score indicating general artistic tendencies
(71). also, using a questionaire as a criterion, Vaughn and Krug
(70) found that Sum C correlate -.52 £ .08 with Bernreuter's
:neasure of neurotic tendency. Correlating CF and the total neuro-
tic score on the doodworth-Mathews Psychoneurotic Inventory, Hertz
(28, 29) obtained a correlation of -.578 f .032 and concluded that

the greater the emotional instability, the more neurotic involve-

ment, and the less adaptability, the larger the CF score. At the ;
same time she found more FC in subjects who obtained low scores on
this inventory.

These correlational studies aypear to support the validity
of the Rorschach in regard to color reaponses but it must be re-
membered that the validity of the criterion measures used remains
open to question.

Genetic and Normative Studies. Rorschach's tentative norms

 

were based on L05 men and women. For his normal subjects he re-

ported averages of 3 FC, 1 CF, and O C re3ponses. Klopfer and

 

 

 

 

Margulies (39) studied children aded two through six years and

found that by the age of six, FC dominated the tyres of color re—
sponses given as compared to earlier ages at which C or CF res;on~
335 were predominant.

bsing a group of children aged three through seven years
Ford (27) obtained reSnlts which were somewhat divergent from those
reported by Klopfer and Largulies (39). Ford found that at the
age of six and seven years CF was the dominant factor whereas KloP-
for and Margulies reported that PC was dominant by the age of six.
Beck reports, on the basis of the Rorschachs of 131 children, aged
six through fourteen years,

"Pure color declines but color-form and form—color hold
steady into the adolescent years. In psycholoyic language,
with learning, with acquisition of character, and of the adap-
tive personality, uninhibited emotiunal response diminishes
and regard for environment is retained or increases...The per-
sonality groups that corresyond with C and CF are the not yet
nature, the socially maladjusted and the regressed” (9, p. 97).

Anes, et. al., (1) made a study of childhood Rorschach re-

sponses using a sample consisting of fifty records at each half-
.year level from two to five and one-half years and at yearly inter-
vals from six to ten years (650 records in all) and concluded,

" he overall trend, though variable, is for there to be a relat-
ive increase in PC reagonses, a decrease in CF and C." One aspect
(If this study which deserves attention, however, is that the sam-
‘ple cannot be considered as representative of the general populat-

icni. Over three-fourths of the children cane from above averabe

scxxio-economic levels and a sinilar proportion were above average

55
in intelligence.

Hertz and Baker (31) studied the color responses of adoles-
cent boys and girls and provided tentative norms for these groups
which indicate a gradual trend toward more FC and less C and CF
between the ages of l: and 15 years. KlOpfer and Kelley (38) in-
dicate that the normal adult should give some color responses but
the sum C and CF should not exceed Sum PC. In contradiction to
t is, however, Beck, et. al., (10) found, in a normative study on
adults, that CF predominated over FC.

On the whole it can be said that the results of genetic and
normative studies on the Rorschach seen to indicate a general
tendency for there to oe an increase in FC and a corr.eponding de~
crease in C and CF resoonses as a concomitant to increasing age
from Childlluod to adulthood.

Experimental Studies. Experimental studies on the validity
of the assumptions regarding color reSponses on the Rorschach have
been few. Williams (75) found only a small and questionably signi-
ficant relationship between the form-color integration ratio and
efficiency of behavior during stress but found a hig ly signifi-
cant correlation betwcen this factor and F#%. Establishing a bas-
al performance level on the digit-symbol test he measured the de-
crement in performance which occurred under the stress of physical
discomfort, noise, critical observers, etc. The limited number
(N 25) and type of subjects used may have been limitations of the

study but as a pioneer investigation it showed the fruitfulness of

 

 

56'

the application of eXyerinental methods to such problems.

Studying the properties of the Cards themselves, Lazarus
(Al) compared performances on standard and achromatic blots and
found no differences in regard to ”color shock.” Sappenfield and
Buher (58) used the same technicue to test the assumption that
productivity on cards 8-9-10 was related to the color value of the
cards. They found no significant differenCes but it is well to
remember that both of these studies used the group Rorschach pro-
cedure. In view of the present state of knowledge it is cuestion-
able whether such results are directly comparable to those obtain-
ed with the stendardly administered Rorschach. Canter, however,
administered chromatic and achromatic versions of the test indi-

dually and also found negative results in regard to "Color

F.

v
shock" (20).

Siipcla (be) investigated the effects of color upon her-
sehach responses and found that color alone did not appear to pro-
duce affective disturbance but that color-form incongruity did.
That is, the blots which produced such disturbances were those in
which the color was incongruous with the usual forms seen in the
achromatic series of blots.

Fitzgerald (26) attempted to relate Rorschach performance
to social behavior by correlating the form—color integration ra-
tio with ratinés of social adaptability in school children and
young adults and found a significant relationship in the predict-

ed direction.

In general, the few enperimental studics which have been
reported on the validity of the color responses have produced re-
Sults which are inconclusive. Nevertheless, the results indicate
that the experinental method holds forth promise as a means of
dealing with the problem.

Summary. In general, the problem of the validity of the
assumptions regardinQ the Rorschach colcr responses remains un-
settled. Benton (lb) after reviewing the studies on color re—
s,onses Concludes that while ezhwirical studies show. that the FC
response is cowrelated positively with EmotiOnal stability, those
studies attexgting to show the validity of the assumptions by
manipulating tne color factor have been unsuccessful. Perhaps
Beck's statement regarding the [roblem of demonstrating the valid-
ity of these assumptions best summarises tLE situation; the state-
ment is, "The ones (factors) which are having the longest wait

are M and C, these elusive indicies to the unConsci us wish-living

and to the emotions. Their day will COJU too" (ll, p. 772).

 

 

 

m
V

 

fifCZD E1 @315:

 

21A@%

58

g C
C?

 

 

 

 

fiib

A
La

 

1

,f‘L,
LU—

 

A

 

 

 

L:

if? D O A

‘7
"th 0 A :37 A O i? A [:1

 

i? D £135 A . E735 51‘?

ACES)

 

 

 

 

 

 

 

 

 

Di} 7:? A d}. A A L! Q

71?.
CCbA

 

 

'Oik E] {'2 A A o

59

APPENDIX D

PHOTOGRAPH OF THh MIRROR-TRACING APPARATUSl

 

1 In order to allow a view of the design the card-
board used to prevent direct vision is not shown.

 

3.

9.

10.

12.

13.

60
REFERENCES

Ames, Louise B., Learned, Janet, Metraux, Ruth & Walker, R. N.
Child Rorschach responses: developmental trends from two to
ten veer . New York: Paul Hoeber Inc., 1952.

 

 

Anderson, J. E. Changes in emotional reaponse with age. In
Feelings and emotions by Reymert, M. L. (Ed.). New York:
McGraw-Hill, 1950.

Baker, L. M. & Harris, Jane S. The validation of Rorschach
test results against laboratory behavior. J, Clin. Psychol., g
19h9. 5, 161-164. 1

 

Barker, R. Frustration as an experimental problem. V. The f
effect of frustration upon the cognitive ability. Character
are 2212-. 1938. 7, lbs-150. m
Beek, 8. J. The Rorschach test as applied to a feeble-minded .4.

group. arch. g; Fszchol., 1932, No. 136, 8A.

Beck, 3. J. Personality structure in schizophrenia: a Rorschach
investigation in 81 patients and 64 controls. Nerv. é'hent.
Dis. Monog§., 1938, No. 63.

Beck, 8. J. Rorschach's test. Vol. I. Basic processes. New
York: Grune & Stratton, 19LA.

Beck, 8. J. Rorschach's test. Vol. 1;. §.varietz 22 personal—
221 pictures. New York: Grune & Stratton, 1947.

Beck, 3. J. Emotional eXperience as a necessary constituent
in knowing. In Feelings and emotions by Reymert, M. L. (Ed.).
New York: McGraw-Hill, 1950.

Beck, S. J., Rabin, A. 1., Thiesen, w. G., Molish, H. and
Thetford, W. N. The normal personality as projected by the
Rorschach test. ‘J. Psychol., 1950, 30, 2h1—298.

Beck, 3. J. The experimental validation of the Rorschach test.
IV. Discussion and critical evaluation. Amer. J. Ortho-
psxchiat., 1952, 22, 771-775.

Benjamin, J. D. &‘Ebaugh, F. G. The diagnostic validity of the
Rorschach test. Amer.|i. Pszchiat., 1938, 94, 1163-1178.

Benton, A. L. The experimental validation of the Rorschach
test. Brit. J, fled. Pszch01., 1950, 23, h5-58.

 

 

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

2h«

25.

26.

61

Benton, A. L. The eXperimental validation of the Rorschach
test. II. The significance of Rorschach color responses.
Amer. J, Orthopsychiat., 1952, 22, 755-763.

Bergmann, M. 8., Graham, H., & Leavitt, H. C. Rorschach
explanation of consecutive hypnotic age-level regressions.
Psychosom.lggg., 19h7, 9, 20-28.

Boss, M. Psychologisch-charakteroligische untersuchungen bei

anti-sozialen psychopathen mit hilfe des Rorschachsehen 3
formdeutversuches. Zsch. f. g. gs . Neur. g. Psychiat., q

1931, 133, 544-575. Cited in Hertz, Marguerite & Elizabeth ‘
Baker, op. cit.

Boynton, P. &.Walsworth, B. Emotionality test scores of
delinquent and non~delinquent girls. J, abnorm. soc.

Psychol., 19h3, 38, 87-92.

 

Brosin, H. W. & Fromm, Erika. Rorschach and color blindness.
Rorschach Reg. Exch., 1940, A, 39-70.

Brewer, D. The relation between certain Rorschach factors
and cardiovascular activity before and after visqumotor

conflict. 1, Egg. Psychol., 1947, 37, 93-95.

Canter, A. An investigation of the psychological significance
of reactions to color on the Rorschach and other tests.
Thesis, State University of Iowa, 1950.

Counts, R. M. & Mensh. I. N. Personality characteristics in
hypnotically induced hostility. g, clin. Psychol., 1950,
6) 325-330 0

Cronbach, L. J. Statistical methods applied to Rorschach
scores: A review. Psychol. Bull., 19h9, A6, 393-429-

Dubitscher, F. Der'Rorschachsche formdeutversuch bei er-
wachsenen psychopathen sowie psychOpathischen und schwach-
sinningen kindern. Zsch. g. Q. ggg. Hear. 3, Psychiat.,
1932, 1A2, 129-158. Cited in Hertz, Marguerite &.E1izabeth
Baker, op. cit.

Edwards, A. L. Egperimental design in psychological research.
New York: Rinehart, 1950.

Eichler, R. M. Experimental stress and alleged Rorschach
indices of anxiety. J. abnorm. soc. Psychol., 1951, 46,

Fitderald, J. C. A study of Rorschach color factors as in-
dicators of social acceptance. Thesis, State University of
Iowa, 19h9.

2.7 .

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

62

Ford, nary. The application of the Rorschach test to young
children. Univer. Minn. Child Wolf. Monogr., 1946, No. 23.

Hertz, Marguerite. Tho norschach ink-blot test. Thesis,
Western Reserve University, 1932.

Hertz, Marguerite. The Rorschach ink-blot test: Historical
summary. Psychol. Bull., 1935, 32, 33-66.

Hertz, Marguerite. The validity of the Rorschach method.
Amer. J, Orthopsyghiat., 1941, 11, 512-550.

vol?"

Hertz, Marguerite & Elizabeth Baker. Personality patterns in
adolescence as portrayed by the Rorschach ink-blot method:
II. The color factors. J, ggg. Psychol., 1943, 28, 3-61.

Ingram, Winnifred. Prediction of aggression from tho Rorschach
test. Thesis, Northwestern University, 1950.

Jorsild, A. T. Child Psychology. (3rd. ed.) New York:
Prentice—Hall, 1947.

Kardiner, A. The individual and his society. New York:
Columbia University Press, 1939.

Kelley, D. M. & Levine, K. Rorschach studies during sodium
amytal narceses (abstract). Rorschach Res. Exch., 1940,
4, 146.

Kelley, D. M., Margulies, Helen, & Barrera, S. E. The
stability of the Rorschach method as demonstrated in
electric convulsive therapy cases. Rorschach Res. Exch.,

1941, 5, 35-43.

Kerr, M. The Rorschach test applied to children. Brit. J.

Klopfer, B. & Kelley, D. M. The Rorschach techniguo. é

manual :25 g projective method 2; personality diagnosis.
'New York: World Book, 1942.

Klopfer, B. & Margulies, Helen. Rorschach reactions in
early childhood. Rorschach Res. Exch., 1941, 5, 1-23.

Lane, Barbara M. A validation test of the Rorschach move—
ment interpretation. amer. J. Orthopsychiat., 1948, 18,

Lazarus, R. The influence of color on the protocol of tho
Rorschach test. J. abnorm. Egg. Psychol., 1949, 44, 506-516.

 

 

1&3.

44.

ASe

46.

47.

[+8o

49.

50.

51.

52.

53.

54.

55.

63

Loosli-Usteri, Mar uerite. Le diagnostic individuel chez
l’enfant au moyen.du test de Rorschach. Paris: Hermann,
1937. Cited in Hertz, Marguerite & Elizabeth Baker, op.cit.

McGeoch, J. A. & Irion, A. L. The psychology 9; human learning.
New York: Longmans, Green, 1952.

McNemar, Q. Psychological statistics. New York: Wiley, 1949.

Nalmo, R. B., Shogass, C., Belanger, D. J. & Smith, A. Motor
control in psychiatric patients under eXperimental stress.
g. abnorm. soc. Psychol., 1951, 46, 539-547.

Masserman, J. H. Principles pg dynamic psychiat_y. Philadelphia:
Saunders, 1946.

Miller, D. R. Responses of psychiatric patients to threat of
failure. g, abnorm. 39g. Psychol., 1951, 46, 378-387.

Morgan, J. B. Child psyghologx. (3rd. Ed.) New York: Farrar
and Rinehart, 1942.

Murphy, G. Personality. A biosocial approach 33 origins and
structure. New York: Harper, 1947.

Oeser, 0. Some experiments on the abstraction of form and
colour. II. Rorschach tests. Brit. 1, Psychol., 1932, 22,
287-3 23 o

Pfister, O. Ergebnisse des Rorschachschen versuches bei
oligOphrenen. Allg. Zach. g. Psychiat., 1925, 82, 198-223.
Cited in Hertz, Marguerite & Elizabeth Baker, op. cit.

Piotrowski, Z. Tentative Rorschach formulae for educational
and vocational guidance in adolescence. Rorschach Res.
EXCl‘lo, 19143, 7, 16-270

Rabin, A. I. Validating and experimental studies with the

' Rorschach method. In An introduction to projective tech-
nigues (Anderson, H. H. & Anderson, Gladys, L. Eds.). New
York: Prentice-Hall, 1951, Pp. 123-146.

Rapaport, D. Diagpostic ppychological testing. 191.‘l_.
Chicago: Yearbook, 1946.

Rickers-Ovsiankina, Maria. The Rorschach test as applied to
normal and schizophrenic subjects. Brit. 1, med. Epychol.,
1938, 17, 227-257.

 

56.

57.

58o

59.

60.

61.

62.

63.

64.

65o

66.

67.

68.

69.

64

Rorschach, H. Psychodiagnostics. 2nd. ed. (English trans-
lation by P. Lemkau & B. Kronenberg) New York: Grune &
Stratton, 1942.

Ruesch, J. & Finesinger, J. E. The relation of the Rorschach
color reaponse to the use of color in drawings. Psychosom.
gag., 1941, 3, 370-388.

Sappenfield, B. R. & Buker, S. L. Validity of the Rorschach
8—9—10 percent as an indicator of reSponsiveness to color.
J. consult. Psychol., 1949, 13, 268-271. 3

‘ uu's'.

Sarbin, T. R. Rorschach patterns under hypnosis. Amer. J,
Orthopsychiat., 1939, 9, 315-318.

Saul, L. J. Emotional Maturity. Egg development and dynamics
2; personality. Philadelphia: J. B. Lippincott, 1947.

Schachtel, L. On color and affect: contributions to an under-
standing of Rorschachs test. 11. Ps chiatr , 1943, 6, 393-409.
f
Schachtel, E. & Hartoch, nnna. The curve of reactions in tho
Rorschach test. A contribution to the theory and practice
of Rorschach's psychodiagnostic ink-blot test. Amer. J.
OrthOpsychiat., 1937, 7, 320-348.

Sherman, H. & Jost, H. Frustration reactions of normal and
neurotic persons. J. Psychol., 1942, 13, 3-19.

Siegel, M. G. The diagnostic_and prognostic validity of the
Rorschach test in a child guidance clinic. Amer. J. Ortho-

psychiat., 1948, 18, 119-133.

Siipola, Elsa. Implicit and partial reversion-errors: a
technioue of measurement and its relation to other measures
of transfer. J. 5&2. Psvchol., 1940, 26, 53-73.

Siipola, Elsa. The influence of color on reactions to ink-
blots. J. Personal., 1950, 18, 358-382.

Syménds, P. M. 239 dynamics 9; human adjustment. New York:
Appleton-Century, 1946.

Symonds, P. M. Dynamic psychology. New York: Appleton-
Century, 1949.

Thornton, G. R. & Guilford, J. P. The reliability and meaning
of erlebnistypus scores in the Rorschach test. J. abnorm.
soc. Psychol., 1936, 31, 324-330.

1| nlllllbll, I‘er' .4
- . .

70.

73.

74.

75.

76.

77.

78.

65

Vaughn, J. & Othilda Krug. The analytic character of the
Rorschach ink-blot test. amer. J. Orthopsychiat., 1938,
8 , 220-229 e

Vernon, P. s. The Rorschach ink-blot test. Brit. J. Egg.
Psychol., 1933, 13, 89-118, 179-200, 271-291.

Vernon, P. E. Recent work on the Rorschach test. J. Ment.
Sci., 1935, 81, 894-920.

 

Wallen, R. The nature of color shock. J. abnorm. goo. Psychol., 1

Wilkins, N. L. & Adams, A. J. The use of the Rorschach under
hypnosis and under sodium amytal in’military psychiatry.
J. ggg. Psychol., 1947, 36, 131-138.

Williams, M. An experimental study of intellectual control
under stress and associated Rorschach factors. J. consult.

Williams, M. The eXperimental validation of the Rorschach
test. 1. Experimental correlations. Amer. J. Orthopsychiat.,

1952, 22, 749-7515 o

Young, R. A. & Higgenbotham, S. A. Behavior checks on the
Rorschach method. Amer. J. Orthopsychiat., 1942, 12, 87-94.

Zander, A. F. A study of experimental frustration. Psychol.
Mono r., 56, No. 3, Whole No. 256., 1944.

A

 

‘I.Y~l\~7

-- . 1"7"Fr!" ,~’
l‘tlL;‘.-L;u1..‘u\l L‘Unl‘

JAN3 1 '55

ROHM USE GNU
mm

 

 

 

"'TITI'I‘WL‘IfllLfiflflflTHflfljlflflfififlflflflfljflflfl‘l’ffs