University of Nebraska at Omaha
DigitalCommons@UNO S#$ W
12-1986
Roberts' Apperception Test for Children: Referred and nonreferred student proles Evelyn Harm Headen University of Nebraska at Omaha
F**4 0 # ##* 40 : ://#&*++0.+.$#/0#$4 P % $ P05*&5 C++0 R$++$#$# C H$#$, E3$*5 H+, "R!$0' A$$ T$0 T$0 % C*#$: R$%$$# # $%$$# $%$$# 0#$ *$0" (1986). Student Work. P$ 269.
70 7$00 0 !& 5 % %$$ # $ $00 !5 D&*C++0@UNO. D&*C++0@UNO. I 0 !$$ $$# % *0 S#$ W !5 $# #+0 % D&*C++0@UNO D&*C++0@UNO.. F +$ %+, *$0$ #&*++0@+.$# #&*++0@+.$#..
ROBERTS' APPERCEPTION TEST FOR CHILDREN: REFERRED AND NONRE FERRED STUDENTS' STUDENTS'
PROFILES
A Thes Th es is Presented to the Department of Psychology and the Faculty of the Graduate College University of Nebraska
In Partial Fulfillment of the Requirements for the Degree Specialist in Education University of Nebraska at Omaha
by Evelyn Harm Headen Dec embe r 19 86
ROBERTS' APPERCEPTION TEST FOR CHILDREN: REFERRED AND NONRE FERRED STUDENTS' STUDENTS'
PROFILES
A Thes Th es is Presented to the Department of Psychology and the Faculty of the Graduate College University of Nebraska
In Partial Fulfillment of the Requirements for the Degree Specialist in Education University of Nebraska at Omaha
by Evelyn Harm Headen Dec embe r 19 86
LJMI Number Num ber:: EP72901 EP72 901
All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there ther e are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion.
Dissertation Publishing
UMI EP72901 Published by ProQuest LLC (2015). Copyrigh t in the Dissertation held by the Author. Author. Microform Edition © ProQuest LLC. All rights reserved. This work wor k is protected protecte d again st unauthorized copying under Title 17, United States Code
ProQuest’ —
ProQuest LLC. 789 East Eisenhower Parkway P.O. Bux 1346 Ann Arbor, Arbor , Ml 4 8 1 0 6 - 1346
THESIS ACCEPTANCE Ac ce pt ed for the fa cu lt y of the Gr ad ua te College , University of Nebraska, in partial fulfillment of the requirements for the degree Specialist in Education, University of Nebraska at Omaha.
Committee
Date
r - 2-9 - / ? y &
AC KNO WLE DG EM ENT S
I would like to thank Dr. Norman Hamm and Dr. Thomas Lorsbach for serving on my thesis committee, and Dr. Richard Wi ko ff
for se rvin g as co mm it te e chairma n.
Both Dr. Ha mm and
Dr. Wikoff have been instrumental in seeing me through the School Psychology program and I appreciate their guidance and enc our agem ent over the past several years.
Dr.
Lorsba ch
also deserves a word of special thanks for agreeing to serve on my thesis committee for a second time. Ap pr ec ia ti on
is ex te nd ed to the school di st ri ct whic h
allowed me to carry out my research, to their parents and students for their cooperation, and to my colleagues who offered encoura gement and additional protocols.
Special
thanks to Kate Dodge and Dr. Natalie Gendler for scoring a subset of protocols to establish interrater reliability. Finally,
a thank you to my daug hters Emily, Ka tie and
Beth for bearing with me through one last project, and especially to Dick, my best printer, critic, and supporter.
TABLE OF CONTENTS
Page Ac kn ow led gem en ts List of Tables Ab st ra ct Chapter I
........................................
........................................
...............................................
Introduction ... Statement of the Problem Review of Relevant Literature. . . . Projective Tests C hi ld re n' s A p pe r ce pt i on T es t . . . Roberts Apperception Test for C h i l d r e n 15 P urp ose of the S t u d y 21 ...........................
..........
..................
iii v vi 1 1 2 2 8
..............
II
..........
M e t h o d 24 Subjects 24 Ma te ri al s and Pr oc ed ur es Scoring and Reliability . . . . . . .
25 28
Results
30
............................
.......................
..........
III IV
...........................
Discussion Discussion of Results Clinical Validity Conclusion
........................
40
...............
....................
.....................
References Ap pe nd ix A Ap pe nd ix B
............................................
............................................
............................................
iv
40 44
49 51 58 63
LIST OF TABLES Table 1
Page Mean and Standard Deviat ion of Scale Scores for Referred and Nonreferred Groups
31
Differen ces in Mean Scale Scores using Pairwise Contrasts (Tukey's WSD Criterion)
32
Mean Scores and Standard Deviatio ns for Ad ap tiv e and Cl in ic al Scal es by Group
34
Pearson's Corre lation Coeffic ients for RATC Scales and Selected Subject Variables (Whole Group)
36
Correlat ions Between RATC Scales and Selected Subject Variables for Referred Group
37
Correla tions Between RATC Scales and Selected Subject Variables for Nonreferred Group
38
....................................
2
.............................
3
.................................
4
...........................
5
....................................
6
........................
v
AB STR AC T
The present study examined a recently published projective test,
the Roberts App ercep tion
Mc Ar thu r & Roberts,
1982).
Test for Childre n
The su bj ec ts we re
(RATC;
58 students,
grades 18, from a midwestern suburban school district. Onehalf of the students had been referred to the school ps yc ho lo gi st for an evaluati on, pr ob le ms
ei th er be ca us e of be ha vi or al
or a su sp ec te d le ar ni ng disa bili ty ,
threey ear reevalu ations
(as required
or for ro ut ine
by law).
The second
group of students were randomly drawn except for the stipulation that they not have been previously referred for a psycho logica l
evaluation.
All subjects were given the
RATC; in addition, the majority of the referred students had scores available for the Wechsler Intelligence Scale for Children Revised and for an individually administered achie vement
test battery.
For the nonrefer red
group of
children, California Achievement Test scores were recorded wh ene ver they wer e available.
An al ys is
of the data
show ed
that referred and nonreferred groups of students differed only in the kinds of resolutions they provided for identified problem situations, with the former group tending to give quick, e asy answers resolution).
Nonreferred
(the lowest level of
students gave proportion ately more
resolutions to stories.
RATC scales may be combined in a
global fashion into Adaptive and Clinical scores; contrary to expectation, groups did not differ in their mean scores on this measure.
Correla tions between the RATC scales and
other measures such as IQ and achievement were also examined.
The val idi ty of the RATC as a clini cal tool was
considered.
It appears
to offer some advantages
over
similar projective tests which have been used with children, and contributes to a better understanding of a child's emot ional
status.
The findings
of this
study poin t to the
need to teach problemsolving skills to referred children.
1 Chapter I
Introduction Statement of the Problem Projective tests of one kind or another have been employed for many years, if "projective" material is understood to mean that which transcends the obvious content of a stimulus.
Proj ecti ve tech niqu es
have in commo n the use
of unstructured,, ambiguous material which allows the subject relative freedom in responding; they encourage a variety of responses with little subject awareness of the purpose of the test, and interpretation is holistic and mu lt id im en si on al
(Rabin,
1968).
Th em at ic ap pe rc ep ti on tests
form one subset of projecti ve techniques.
These typical ly
comprise a series of pictures, for which the subject is asked to provide a story with beginning, middle, and end, as we ll as a des cr ip ti on of wh at th e ch ar ac te rs are feeling.
The The mati c App erc ept ion Test,
or TAT
sa yi ng and (Murray,
1943) and the Children's Apperception Test, or CAT (Beliak & Beliak,
1949)
are two wel lk now n examples.
pu bl is he d instrume nt,
A rece ntly
the Ro be rt s Apper ce pt ion Test for
Children, or RATC (McArthur & Roberts, 1982) was the focus of this study.
In the prac tice of school psychology,
it is
often desirable to obtain idiographic as well as nomothetic data.
The present research sought to determin e wheth er RATC
pr of il es wo ul d di ff er be tw ee n grou ps of
"normal11 ch ild re n
and those referred for psychological evaluations through the schools. Review of Relevant Literature Projective te st s.
Modern proje ctive techniq ues traGe
thei r begi nnin gs to stimul i such as inkblots. pu bli cat io n reg ar di ng inkb lo ts
ap pe ar ed
in Ge rm any in 1857,
pr es aging Ro rs ch ac h' s Psychodiagnostik, in 1921
(Kl opfer & Davi dson ,
1962).
The first
whic h wa s pu bl is he d
In 1895, Bine t and
Henri suggested using inkblots to investigate visual imagination.
Stories told to pictures,
a less popular
device, was used by Binet and Simon in 1905 as part of their assessm ent of intelle ctual development. period,
howev er,
attempted.
During that time
pers onal ity as se ss me nt as such wa s not
Indeed,
personal ity theory was
just begin ning to
develop in the 1920s in conjunction with psychoanalysis, and with it the fi eld of pro jec tiv e techni ques .
It wa s Mur ray,
under whose leadership the Thematic Apperception Test (TAT) was develo ped, 1938
(Rabin,
wh o in tr od uc ed the term
"p roje ct ive tests"
in
1968).
Projective tests have enjoyed considerable popularity, judging by the number of tests listed in the "Personality" section in the Mental Measur ements
Ye ar bo ok .
Acco rdin g to
Rabin (1968), the growth of projective measures in the 30s and 40s paralleled the increasing importance of clinical
3 ps yc ho lo gi st s, effort.
pa rt icu larl y in co nn ec ti on wi th the wa r
Interest grew
pu re ly ps yc hom etr ic
in instrument s
information.
that offered more
Fr ank
than
(1939) ob se rv ed that
we may see in di vi du al s as or ga ni sm s in the wo rl d of nature, as membe rs of a group,
or in private worlds dev eloped under
the impact of experience .
It is this priv ate wo rld
that
clinicians have tried to tap with projective measures. Wh il e st an da rd iz ed test s pr ov id e in fo rm at io n abou t pe rs on s in relation to norms, they do not help in discovering individual
characteristics.
Furthermore,
while peopl e can
conceal this information in selfreport or inventory types of measures, the indirect approach of projective measures is mo re like ly to pr ov id e sp on ta ne ou s re fl ec ti on s of pe rs on al it y,
if on ly be ca us e cl ient s do n't know wh ich is the
best, foot to pu t fo rw ar d
(Zubin,
Eron,
& Schumer,
1965).
Rabin (1960) suggests that "externalization" might be a mo re ap pr op ri at e te rm than pr oj ec tion , refers to a Freudi an d efe nse mecha nism.
since the
latter
1 As he poin ts out,
however, projection may be seen as a continuum; we are interested in the portion which is described as the private world.
The cl in ic al aim is to ev al ua te the total person,
including an assessment of ego, ego strength and defenses, assets
and liabilities.
Rabin
feels that the freedom with
which a subjec t can ma ke be li ev e or tell a story
(as op po se d
to factual description) is an index of the ego's freedom
4 from threat.
Other nonpa thol ogica l aspects of the
per so nali ty are of in te re st to the cl in ic ia n as we ll of coping with reality,
identity,
w ay s
interpers onal
relationships, problemsolving style, and so on. Responses to projective stimuli are determined by three classes of stimuli: the stimulus properties of the test; the ba ck gr oun d
(both ph ys ic al and ps yc ho lo gi ca l)
of the test
situation; and personality characteristics of the subject (both org anis mic and acquired)
(Murstein,
1968).
Und erl yin g
pr ojec tiv e techn iq ue s is the as su mp ti on that re sp on se s are determined and predictable, rather than accidental. Evaluation further assumes that the protocol is an extensive enough sampling of personality to allow judgments about it; that psychological determinants are basic and general; and that it taps the "durable essence" equally in different subjects.
Thus the products
of proj ectiv e techniqu es may
not always predict behavior, but they do help interpret observed behavior (Zubin et al., 1965). Klopfer (1968) has pointed out that projective techniques are necessary but not sufficient for understanding both individual personality and group characteristics.
He suggests a cont inuum of measures,
exten ding from what is essentiall y a structured in tervie w (e.g., sentence completion tasks) to thematic types (such as the TAT).
Mat eria l in the latter is wha t he calls
5 "preconscious," in that the subject is not entirely unaware of the material.
While he calls for more resea rch on
stimulus demand, Klopfer believes that TATtype tests enable the clinical psychologist to gather information about mo tiv ati on
in a wa y that is not av ai lab le th ro ug h pub lic or
selfreports. Wh et he r or not the in fo rm at io n gl ea ne d from pr oj ec tiv e me as ur es Ne ur in ge r
is valid or re li ab le is still a m at ter of deb ate. (1968)
clai ms that the ar ray of th em at ic me tho ds
has outstri pped the amount of respo nsib le research.
The
number of methods "is sometimes disconcerting, and when accompanied by meager reliability and validity information, creates a poor impression of the status of projective tech niques "
(p. 254).
He calls for rep res ent ati ve
standardizing samples, validity research, and extensive normative data from various populations, with constant reeval uations
and replications.
As it now stands,
the
efficacy of such tests depends a great deal on the experience and clinical skills of the examiner; on this po in t vi rt ua ll y all so urces agree. the prolifera tion of techniques
Ra bi n
(1960)
feels that
is not the problem,
so much
as the lack of their relationship to systematic personality theories.
On the other
hand,
as he point s
out,
these
met ho ds are of te n ju st ifie d in the long run because, pr ac ti ca l instrume nts,
th ey st im ul at e theory.
as
Zubin et al.
(1965) feel that predictive,
concurren t,
and construct validity have been demonstrated for the TAT, but be li ev e the ev id en ce is not yet in re ga rd in g co nt en t validity.
Klopfer
(1968) considers predi ctiv e effic iency
mo re val ua bl e as a cr it er io n of va lidi ty th an co nc ur re nc e be tw ee n tests,
no ti ng that Ro rs ch ach cont en t pr edic ts
observ able behavi or better than formal scoring.
The scoring
system used also determines the findings to some extent. For'example, Beliak (1968) suggested that Witherspoon (1968) failed to find Oedipal conflict in CAT protocols because he didn 't code corr ect ly for it.
Mur ste in
(1968)
found that a
subject's selfconcept was more important in determining a TAT response than an objective assessment of the possession of the trait
(in his experime nt,
hostili ty).
However,
the
stimulus properties of the cards outweighed all other determinants; Murstein’s moral is that conducting content analysis without considering the stimulus impact is inappropriate. At the risk of ov er ge ne ra li zi ng ,
it ma y be safe to say
that the issues of reliability and validity have not yet be en se tt le d sa ti sf ac to ril y for the em pi ri ca l ps yc ho lo gi st , wh er eas the cl in ic ia n isn't much bo th er ed by th em in any case.
Ana stas i
(1976),
pro jec tiv e te ch ni qu es
for example,
suggest s that
shou ld mo re app ro pri at ely be
considered as clinical tools than as psychometric tests.
7 She refers to them as "wideband" procedures, that is, techniques which yield a wide range of information at the cost of a lowered
level of depen dabi lity .
Similarly,
Hertz
(1970) reminds us that as objective data have their subjective aspects, pr ed isp os it io ns
the reverse
is also true.
Individual
inter ac t in a co mp le x man ne r wi th the social
and situational milieu.
She cautions against an
overreliance on computerized scoring and other actuarial met ho ds be ca us e th ey do not ade qua tel y ac co un t for environmental influences, nor allow for clinical experience. Errors may also be introduced in the gathering and ma nipul atio n of data.
Wh il e a cl in ic ia n' s de ci si on s should
never be based upon a single piece of data, the results of pro je ct ive te ch ni qu es may ne ve rt he le ss
su gges t leads for
further exploration. Many in ve st ig at or s ha ve wa rn ed ab ou t the pr ob le ms involved in using proje ctive
techniques with
children.
It
is necessary to understand normal developmental levels and pr oc es se s in or de r to re co gn iz e de vi an t be ha vi or 1980).
(Magnussen,
Since childr en are in a fluid state of ego
development,
some functions may appear age inappropriate
compared to others.
Age norms mask muc h unevenne ss in
individuals, and performance may be spread over several age levels
(Rabin, 1960).
Studie s wit h you ng chil dren are also
subject to differences because of the verbal nature of the
8 resp ons es
(Murstein,
1970).
Rabi n point s out that tests
like the CAT require a fairly persistent and continuous set, and that older children, with sturdier ego structures, are mo re pr od uc ti ve and reveal in g.
For ve ry yo un g children ,
he
recommends the least restrictive methods which allow maximum freedom and spontaneity, such as free art, play, and puppet ry. Haworth (1966), tracing developmental trends from a variety of theoretical perspectives ranging from Freud to Piaget, concludes that there is a general progression from a diffuse, global state, through phases of differentiation, to a final complex but smooth integration of the organism. prac ti ce, prot oc ol s,
thi s me an s we for ex amp le,
sh ould ex pe ct yo un g ch il dr en 's to re fl ec t un rel at ed
In CA T
fa nt as ie s or,
if they focus on the stimulus, descriptive responses without regard to nuances
of thought
or feeling.
Primary
children
wo uld be ex pe ct ed to be mo re cr ea ti ve but wi th an aw ar en es s of reality, while the older child would be put off by the childishness of the pictures. Ch il dr en 's Appe rcep tion
Test
(C A T ).
The original
pu rp os e of the CAT, whi ch is cle ar ly a der iv at iv e of the TAT, was to provide situations more specific to typical chi ldh ood problems
(Beliak & Beliak,. 1959).
The test is
intended for use with children aged three to ten, of both sexes and all ethnic groups.
Animals were chosen for the
9 original series because it was believed that children would identify more freely with animals; problems of cultural or racial identification were also avoided in this way (although it has been pointed out that the furnishings and details such as pipe and cane are not culturefree). Beliak (1954) feels that children will be very pr oduc tiv e in th ei r stor ies be ca us e th ey use mo re projec ti on ,
symbol
and wi sh fu lf il lm en t th em es are ma ni fes t and
readil y given.
He does warn that, rather than showing the
entire character structure, children's themes may reflect transitory problems or developmental stages; what may seem pa th olo gi cal
in an ad ul t may be ap pro pri at e to a ch il d' s
stage of development.
No norma tive data is provided,
however, beyond a simple checklist of typical responses (Beliak & Adelman, 1960). A ten y ea r lo ng it ud in al
st udy us in g the CA T wa s carr ied
out by Witherspoon (1968) with subjects ranging in age from 2:9 to 6:5 at the beg inn ing of the study. ba se d on the sc or ing of 268 prot ocols,
His concl usions ,
we re as fol lows:
responses were largely apperceptive, with the frequency of nona pperc epti ve responses minimal by age eight;
sex
differences were virtually nonexistent, except possibly at ages
3 and 4;
judged by fre que ncy and intensity,
the
dynamics of parent identification, aggression, and orality we re be st ex pl or ed by the CAT;
fears,
si bl in g rivalry,
10 Oedipal concerns, toileting and cleanliness, and sexuality we re rel at iv el y in fr eq ue nt responses.
Wit he rs poo n also
cited a study by Rosenblatt (1958), in which 400 stories of children aged 3 to 10 were examined from a psychoanalytical po int of view.
Fe w uniform age tr en ds we re
found,
and
statistically significant differences were a function of a card's stimulus value. An ot he r st udy led to the co nc lu si on
that ne ed s with the
greatest frequency and intensity had an inverse relationship to the rank order of manifest needs as determined by teacher questio naires
(Sanford,
pr ed at es the CAT,
1943).
Alth ough this resear ch
the re su lts are su gg es ti ve for
appe rcepti on tests in general.
(Sixty differ ent
stimulus
cards were used, from the series used at the Harvard Psychol ogical
Clinic.)
The needs most
freely expres sed in
fantasy were those least frequently shown in overt behavior, namel y aggression,
acquisition,
and autonomy.
The author
suggests that perhaps these were the needs most commonly inhib ited by the children 's culture.
Becaus e the
correlations between covert and manifest expression of needs was
so low
(.11 ov er all),
are no indicator
she co nc lu de s that fa nt as y scores
of overt behavior.
"Good" outcomes were
given by 67% of the subjects; for these children, positive correlations were found with n af fil iat ion , deference, nurtu rance ,
and play.
In general,
then, happ y chil dren
11 tended to tell pleasant stories. Hawo rth
(1966),
in a revi ew of the nor mat ive
literature, noted that normative responses are the ones that could be predicted by simply describing the pictures. Useful responses, clinically speaking, are those that deviate from the stimulus, introduce additional figures or objects, or give unusual interpretations to the stimuli. She feels that further tabulational studies are unnecessary, and tha t more is to be gai ned from e xam ini ng th e meani ngs and implicat ions
of the unusual responses.
und ers cor ed by Zubin et al.
This belief
is
(1965), who cons ider devia nt
responses a reflection of possible pathology. Mu ch re se ar ch wa s di re ct ed at the eff ic acy of an im al versus human figures in the CAT stimulus cards.
Reviews of
the literature (Beliak & Hurvich, 1965; Murstein, 1970; Neurin ger,
1968)
ag re e that the ove rw hel mi ng nu mb er of
studies failed to show superiority of animal over human pi ct ures,
al th ou gh Be li ak argu es
exa ctl y equiva len t
(e.g.,
that the forms we re not
in Card 10 the puppy could be seen
as being brushed, where as in the human version, mo re
spanking is
like ly to be elic ited). Wh il e Be li ak be li ev es
that st ru ct ur ed
stimuli vi ol at e
the principle of projective testing, Murstein maintains that there is a curvilinear relationship between ambiguity and pr oj ecti on .
In this reg ard,
Ep st ei n
(1966)
re co mm en de d
12 using stimuli with at least three levels of ambiguity (although he was not addressing the CAT in particular). Highly structured stimuli are likely to activate latent drives, whereas highly ambiguous stimuli may not arouse the subj ect at all.
In Epst ein' s view,
TAT stimuli wer e
pa rtic ular ly po or in ev ok in g th em es of pa re nt ch il d relationships,
sex,
and aggression.
Less ambiguou s
stimuli
have the added advantages, he believes, of requiring subjects to deal'with material important to the examiner, and of allowing interpretation of responses in reference to a stimulus of known significance. Whe n as se ss in g the va lu e of the CA T as a cl in ic al instrument, the lack of validity and reliability studies is still problematic.
Holt
(1950) critici zes Beliak's
approach
to testmaking, which is simply to try it and see if it works.
In fa ir ne ss to Belia k, wh at he sai d
(in 1968)
is
that singlecase studies are useful in projective tests, be ca us e over time th ey pro vi de da ta th at can be an aly zed stat istical ly. Wi th er spo on' s
Beli ak (1968)
(1968),
in turn,
crit ici zed
at te mp t to or ga ni ze his lo ng it ud ina l
data by means of factor analysis; Beliak claimed that while it migh t have reliability,
it did not have enough const ruct
validity. At
least two wr it ers
(Neuring er,
1968; Wirt,
the CAT as being most valuable for a play type of
1970)
see
13 apperception test, used to relax the inhibited or repressed child, although Wirt feels structured doll situations to be superior
in that instance.
Wirt
further suggests using
the
CAT Supplement (CATS, a set of ten pictures designed to tap specific concerns such as peer relationships, health, ph ys ic al disabili ty ,
etc.)
to asse ss par tic ul ar problems ,
espe ciall y in the context of play.
The same critici sms
leveled again st the TAT are relevant to the CAT,
for
example, that it is subject to a wide range of distortion (Adcock,
1970).
There
is no reas on wh y basi c pe rso nal ity
factors would not be represented in the stories, but as Ad cock po in ts
out,
th er e is no gu ar an te e that th ey wi ll be.
It can be helpful in picking up specific sources of dist urban ce and for indic ating general interests,
sources of
current motivation, and clues for clinical discussion, but, in his view,
it does not provid e a reliabl e measur e of any
one trait, or a profile of personality traits. Mor ia rit y
(1968)
ca ut io ns ex ami ne rs to re me mbe r the
pu rpo se of the CA T and re ma in me aning of in di vi du al
se ns it iv e to the dy nam ic
responses.
In or de r to un der sta nd an
individual child, one must go beyond the norms to see how pr ob le ms
are pe rc ei ve d and handled,
affec t adjustment.
and fur the r,
ho w they
In a longter m follow up of children
assessed as preschoolers, Moriarty noted that coping mec hani sms us ed by ch il dr en to re du ce
st ress and cl ar if y
14 demands wer e highly personal;
the effect, effect,
per p er si s te d into in to late la te ad ol es cenc ce nc e. pr ob le ms mi ss es
if not the for form, m,
Si mp le en um er at i on of
the th e dy na mi c m e a ni n g fu l n es s w hi c h is the th e
strength of the proje ctiv e method.
Altho ugh there are
common conditions and uniformities, each individual is unique; we must "seek to understand this uniqueness in relation to standard expectations, feelings, conflicts, and resolutions"
(p.
418).
Other instruments have been devised for use with children.
In contrast
to to the lack of prel imina ry resear ch
noted with the CAT, the Michigan Pictures Test (Andrew, Hartwell,
Hutt, Hutt,
1400 children.
& Walton,
1953) 1953) was ini tial ly given to over
The authors
should be realis realistic, tic,
felt that stimulus mater ial
reflecting everyday events with which
childre n could identify. identify.
The test is designed for ages
814, and supposedly evokes more themes of achievement and concern over peer affiliation than does the TAT or Symonds1 Pict ure Stor y Test
(Symonds, (Symonds,
1939) 1939)
(Neuringer, (Neuringer,
1968). 1968).
Only
three variables differentiated between high and low adjusted children; the Tension Index has held up best over cross val idati on studies, studies,
accordin g to Neuringer.
This is
thought to be a global reflection of basic, unresolved needs. needs.
While the well
adjusted
subjects referred more
to to
love and personal adequacy needs, poorly adjusted children referred more to extrapunitive and submission needs.
15 Ma la dj u st ed
ch il dr en us ed mo re past pa st te ns e co ns tr uc ti on s,
contras t to well adjusted, adjusted, who used mor e present tense. tense.
in In
view of the criticisms regarding the psychometric properties of the CAT, Neuringer ironically concludes that the "objectification and precisioning of a projective technique seems to diminish the amount of data from which adequate in fer enc es can be mad e" Symonds'
(p. 236). 236).
(193 (1939) 9) Pict ure Story Test was in tende d as a
TAT for adolescents.
Admi nist ratio n and scoring are
similar, except that students are asked for absurd, wild, or silly stories, which may lead to erroneous interpretations. The pictures tend to be gloomy and dismal, and there are insuffi cient norma tive
table tables. s.
Since it has not
demonstrated any superiority over the TAT, the test is not wi d e l y us ed
(Neuri (Ne uri nge r,
1968). 1968 ).
Ot he r ap pe rc ep ti ve
tests, test s,
such as the Blacky Pictures Test (Blum, 1950) and the Ma k e A P i ct u r e S to r y Te st
(Sc hne idm an, an ,
1949)
are
infrequently used in school settings and thus will not be discussed here. Roberts Apper cept ion Test for Child ren
(R (R A TC ).
Having
explored the history of projective techniques in general, and issues relevant to two wellknown apperceptive instruments, the TAT and the CAT, we will turn now to a new instrume nt,
the RATC
(McArthur & Roberts ,
1982). 1982).
The RATC
grew out of doctoral research on projective stimulus
16 var iat ion
(Roberts, 1958). 1958).
Four types of stimulus
figures
animals, children, adults, and stickmen were presented in four identical situations aggression, social relationship, heterosexual relationship, and an ambiguous situation. Human figures elicited the most significant projective mate ma teri ri al .
Mo st impo im port rtan ant, t,
the th e si gn i fi ca n ce of the th e si tu at io n
variable supported the conclusion that pictures should be selected on the basis of their appropriateness for investigating aspects of personality. The stimulus as determinant of response content was further explored by McArthur (1976), When she compared six equivalent pictures from the TAT, CAT, and RAT (as it was origi nall y designated).
The RAT was found to be superior to
the more traditional projective material in its ability to elicit situationally related nonstereotyped stories with a high level of projection.
Since the obtaine d stories were
consistently situationally related, those stories that were a departure from the norm could reasonably be considered to indicate psych ologic al disturbance.
The TAT, TAT, in comparison,
resulted in a large number of nonsituationally related stories, stories,
sugges ting that the picture s may be inappropr iate
for children.
Certai nly,
as McA rth ur
(1976 (1976) )
and Epst ein
(1966) both emphasize, a prerequisite for the interpretation of projective stories is a mutual understanding of the situation,
shared by the subject and the examiner.
17 Studies of concurrent validity were subsequently carried out which demonstrated the ability of the RAT to me as ure be ha vi or ch an ge fo ll ow in g ps ycho ther apy
(Hersh,
1979) and to differentiate between clinical and nonclinical families
(Muha,
1977).
Additio nal data wer e collec ted and
guidelines generated through continued experimental use of the RAT, culminating in the publication of the RATC in 1982. Ac co rdi ng to the M a n u a l , limitatio ns of previ ous tests, and CAT,
the RA TC ov er ca me the such as the wid ely used TAT
in the fol low ing ways:
1. The RATC was specifically designed for children ages 6 through 15 and depicts children in all 16 stimulus cards. 2. The RATC emphasizes everyday interpersonal events of conte mporar y life.
It includes
those situations commonly used in thematic pr oj ec tiv e test s relationships,
(e.g., pa re nt chil d
sibling
relationships,
aggression, and mastery) as well as new situations such as parental disagreement, pa re nt al school,
affect io n,
ob se rv ati on of nudit y,
and peer relationships .
The test also
emphasizes the child's ability to cope with situations requiring an appropriately aggres sive response.
Four of the cards depict
18 aggressive situations in varying degrees of explicitness. 3. The RATC stimuli are consistent in their pr es en ta ti on .
All show re al ist ic dr aw in gs of
children and adults executed in a uniform style by the same artist. 4. The RATC employs easily scored, objective me as ure s whic h yi el d hi gh in te rr at er agreement. 5. The RATC provid es no rmativ e data for a sample of 200 welladjusted children ages 6 through 15 to aid in the clinical interpretation of test
results
(p.
1).
Suggested uses for the RATC include clinical assessment and treatment planning, measurement of change, and assessment of developmental changes and situational crises in oth erw ise norma l pr ov ided,
children.
A set of 27 stimulu s
but be ca us e pa ra ll el male/femal e ve rs io ns
cards is oc cu r
for 11 cards, only 16 cards are administered to any given child.
Appen dix A illustra tes the 16 cards
(with male and
female versions interspersed), followed by the authors* description of typical themes and observations regarding the clinical
signif icance
Adap ti ve and Cl in ic al
of each picture. catego ries .
Appe ndix B lists the
The Ma nu al
pr ov id es
detailed procedures for administration, scoring, and
19 interpretation; information regarding development, standardization, and psychometric properties of the test is also included. Psychological evaluation, as practiced in the context of the public schools, typically involves the assessment of intellect ual ability..
Depen ding upon one's theoretical
stance, additional tests may be administered to assess children' s behav ioral
and/or emotional
status.
me as ure s may be "e mp loy ed to as se ss the latter,
A range of from the
Rorschach and projective drawings to sentence completion and selfrep ort inventories.
A techn ique whi ch is applic able
across the age range of school children, is perceived as an enjoyable activity, and which allows for relatively uncensored selfdisclosure is a desirable addition to the ps yc hol og ist 's te st battery.
The RA TC me et s thes e cr it er ia
and offers, in addition, a reasonably clear method of scoring which yields T scores, normed for various age groups. Research on the RATC to date has focused on clinical po pu la ti on s,
and has dem ons tr ate d the abi li ty of this
instrument to differentiate between adjusted and disturbed children
(McArthur & Roberts,
1982).
Well adj usted
children
differed in their mean scores, with adaptive scales higher than clinical scales; this was particularly reflected in their ability to express positive emotions and to provide
resoluti ons
for their
stories.
original expectations,
Con trary
to the authors'
"reliance on others" and "limit
setting" were found more frequently in the protocols of we ll adju sted
ch ildren.
Go in g to appr opr ia te pe rs on s for
help is adaptive, after all, as is appropriate punishment which
implies the se tt in g of re as on ab le
limits.
Ad jus ted
children responded accurately to stimuli, including both figures and situations; while scores on the clinical side wer e expected,
th ey fell wi thin the ave ra ge ra nge
score s of 4060).
Levels
of proj ectio n,
however,
( T did not
differ for the two groups. Current research on the RATC includes studies with crosscultural and ethnic minority groups, with young children, with children in various types of school settings, and with identified subgroups such as learning disabled children 1985).
(G.
E.
Roberts,
personal
communication,
August
23,
The authors have pointed out the need to explore the
relationship between their standardization sample of "welladjusted" children and a more representative sample of average children, and to look at differences in mean pr of il es
for ch il dr en fr om di ag nos ti c subgroups.
Research in special education has identified several characteristics of learning disabled students, including the following: a lack of selfconfidence and task persistence togeth er with a poor selfco ncept
(Shelton, Ana st op ou lo s, &
21 Linden,
1985);
lower levels of social compete nce and mor e
be ha vio r problems. (McC on au gh y & Ritter,
1986);
and a
dependent learning style and lack of verbal expressiveness (McKinn ey & Speece,
1983).
The pe rce pti ons of par ent s
(Pihl
& McLarnon, 1984) and teachers (McKinney & Feagans, 1983) further confirm the results of selfreport measures and direct observation,
indicating that learning disabled
po pul ati on s ma y be dis ting uis hed fr om norma l ch il dr en on a numbe r of dimensions.
A study whic h employe d the RATC with
learning disabled and normal students found that the Depression and Unresolved scales were consistently higher for the LD group (Wong, unpublished manuscript; G. E. Roberts,
personal
communication,
Aug.
23,
1985).
Similarly, children who are identified as having serious emotional problems demonstrate one or more specific characteristics such as depression, acting out behavior, inability to form appropriate relationships, and so on. Althou gh there are di ve rs e def in it io ns impairment
across
states,
of be ha vi or al
psychologists,
parents,
and
teachers agree that such children are discernible from the normal popula tion
(Cullinan, Epstein,
and McLinden,
1986).
Purpose of the Study The purpose of this study, then, was to compare the pro to co ls of ch il dr en wh o we re ref er re d for ps yc ho lo gi cal services through the public schools with a randomly drawn
22 sample of nonre ferred
students.
Refer red childre n are
frequently found to have a handicapping condition and theref ore to quali fy for special educa tion
services.
A
Mi dw es t su bu rb an scho ol pop ula tio n ma y or may not be ty pi ca l of the pop ula tio n as a whole,
since a numb er of
upw ardl ym obil e, professional families make their homes there.
A previou s study of the
(Jastak & Jastak,
1976),
Wide Rang e Achie vem ent Test
for example,
dem ons tra ted that
local norms were significantly higher than the published norms for primary academic skills (Newville & Kamm, 1984). Therefo re it seemed appropria te test such as the RATC,
to explore local norms
on a
as well.
In consideration of the information presented above, the follow ing question s were
raised:
(a) Will referre d
and
nonreferred groups of children differ significantly on one or more RATC scales?
(b) Will mean Clinical scores be
significantly higher than mean Adaptive scores for the referred group, with either no differences or higher Adap tive scores for the nonr efe rre d grou p?
(c) Be ca us e
Resolution scores do not indicate how many stories a child tells which in fact require resolution, another question arose: Will referred and nonreferred groups differ in the pro por ti on of re so lv ed
st or ies?
(d) Be cau se McA rthu r and
Roberts informally observed that more intelligent children seemed to score higher on Problem Identification, a further
23 question was raised: Will scores on the Problem Identification scale correlate positively with intelligence or wit h standar dized measur es
of achie vement ?
Clinical
Response,
Indicators
(Atypical
(e) While
Maladaptive
the
Outcome,
and Refusal) occurred infrequently in the original research po pu la ti on ,
cl in ic ch il dr en ne ve rt hel ess
me an scores on all In dicators,
ob ta in ed hi gh er
Thus a fina l ques tion:
a similar pattern occur for referred and nonreferred children?
Wi ll
24 Chapter II
Me Ll'iod Subjects The referred group of children included 31 students in grades 18 who were referred for psychological evaluations, or due for threey ear evaluations, school district.
in a midw este rn suburba n
Two protocol s were
eliminated
from the
study, one from a student whose measured intelligence was wi thi n the ed uc ab le mentall y ha nd ic ap pe d
range,
and one from
a behaviorally impaired student who failed to comply with instructions. as follow s: 1315:
The number of students in each age group was
67
N = 10.
yea r olds: N = 5; 89: N = 3; 1012 : Th ere wer e 21 boys
N =11;
and 8 girls.
To form a nonreferred group of subjects, teachers in grades 16, at two elementary schools, were each invited to randomly select five students who had never been referred to the
multi discipli nary team for learning or behavior
problems.
Ex pl anat ory le tt ers and co ns en t forms we re
to thes e famil ies consent,
5 from
(100 in all).
Fro m thos e who
gave
each grade level were rando mly drawn
total of 30 subjects.
One
sent
protoco l was omitted when
fora it was
discovered that the child had been referred for a suspected learnin g disa bilit y at his previous
school.
The number of
nonreferred subjects in each age group was as follows: 67
25 year olds: N 6; 89: N = 11; and 1012: N — 12. we re
14 boys and 15 girls.
Thus ea ch grou p
There
(referred and
nonreferred) consisted of 29 students. Scores on the Wechsler Intelligence Scale for Children Revised (WISCR) were available for 25 referred subjects, with a br eak do wn into Ve rb al subjects. =9);
and Pe rf orm anc e IQ for 2 4
Full Scale IQs ranged from 84113
V erba l IQs,
Pe rf or ma nc e IQs,
78109 781 24
( M ( M
=
92, SD
= 100,
SD
( M
=95,
SD
= 10; and = 12).
Scor es on
the WoodcockJohnson PsychoEducational Battery (WJ), an individually administered achievement test, were also avai lable for 22 of the referre d subjects.
The WJ yields
standard scores in reading, math, written language, and general knowledge. California Achievement Test (CAT) scores were available for all but two of the nonreferred children in grades 36. Total battery scores for those 17 subjects ranged from 5096, M
=68
(normal curve equivalents).
graders who had been tested for
Two second
thedistrict’s Challenge
pr ogram had ob ta in ed pe rc en til escores
in the 90s.
Based
upon their reported reading level at the end of the school year, the other children in grades 1 and 2 appeared to be academically average or above. Ma te ria ls
and Pr oc ed ur es
The 16 pictures of the RATC were individually
26 administered and scored according to standardized instructions provided in the manual. in thei r schools#
Children were tested
in a room apart fro m their classroom s.
The investigator collected all but six protocols, which were obtained from other school psychologists in the same district. Following is a brief description of each of the RATC scales. Ada ptiv e Scales: 1. Reliance on Others (REL) measures the adaptive capacity to use outside help to overcome a problem. 2. SupportOther (SUP0) reflects a tendency to support others by giving assistance, emotional support, or ma te ria l objects. 3. Suppor tChil d
(SUPC) meas ures
sel fsu ffici ency and
matur ity as in di ca te d by ass erti ve nes s or the ex pe ri en ce of positive emotions. 4. Limit Setting (LIM) measures the extent to which pa re nt s or ot he r au tho rit y fi gu re s pl ac e re as on ab le limits on the child in response to a perceived violation of rules or expectations. 5. Problem Identification (PI) indicates the ability to formulate concepts beyond the card and to articulate pr oblem situations. 6. Reso lut io n 1 (RES1)
refl ects a ten den cy to seek easy
27 or unrealistic solutions to problems, including those with a m a g i c a l ,.w ish fulf illi ngf or unrea li st ic quality. 7. Resolution 2 (RES2) indicates a constructive resolution, limited to the present situation. 8. Resolution 3 (RES3) indicates a constructive resolution which goes beyond the immediate problem (e.g., the problemsolving process is fully explained or the solution is generalized to new situations). Clinical Scales': 9. Anxiety (ANX) assesses the frequency of manifest anxiety or apprehension,
including remorse and themes of
illness and death. 10.
Agg res sio n (AGG)
mea sur es the ext ent to whi ch
characters express anger or engage in physic al or verbal aggression. 11. Depression (DEP) used to score stories which contain sadness, despair, or physical symptoms of depression such as fatigue. 12. Rejection (REJ) used to score themes of separation, jealousy, discrimination, or feelings of being left out. 13.
Unr esol ved (UNR)
ind icat es an emot ion al reac tio n left
hangin g or no outc ome to a stated problem. Indicators: 1.
Atypic al Response
(ATY) indicate s extreme deviation
from the usual themes or primary process thinking
28 (includes homicidal or suicidal ideation and child ab us e.) 2. Maladaptive Outcome (MAL) scored when characters make solution to a pro blem more diff icult
(e.g., ru nning
away) or when the story ends with a main character dying. 3. Refusal (REF) scored when a child refuses to give a response, or begins and then stops abruptly. Scoring and Reliability In accord ance wit h the RATC manual,
T scores were
assigned to each raw score for the eight Adaptive and five Clinica l
scales.
(Because Resol ution
3 is expecte d only for
the oldest age group, the 612 year olds have scores for only seven Adapt ive
scales.)
McAr thur and Robert s
con side r T scor es < 40 and > 60
(1982)
(one stan dard dev ia tio n
above and below the mean of 50) to reflect significant differences from a welladjusted population of children. These scores would be equivalent to the 16th and 84th pe rc ent ile ranks,
re sp ec tive ly .
Because the three Indicato rs Ma la dap tive Outcome,
an d Refusal)
(Atypical Response, oc cu rr ed too ra re ly in the
original standardization sample to have the psychometric pro pe rt ie s of sca les ,
cu tof f po in ts wer e ar bitra rily
established by the authors as the raw score at which the cumulative frequency > 90%.
Two experienced school psychologists, who had used the RATC for several years and also attended a workshop given by Dr. R obe rts on the use of the RATC, pr ot oc ol s to es ta bli sh
in te rr at er reli ab il it y.
pr ot oc ol s from ea ch gr ou p randomly selected.
score d a sub set of 10 Fi ve
(referred and no nr ef er re d)
wer e
A reliability figure was calculated for
each scale of each protocol by dividing the smaller score by the larger one.
These were averaged to form mean
reliability figures as follows: Referred group, Adaptive scales = 87%; Referred group, Clinical scales = 92%; Nonr ef er re d group,
Ad apt iv e scales = 84%; Non re fe rr ed group,
Clinical scales = 95%; and overall average for both groups = 89%.
Only two scales with in the nonr eferr ed group's mean
scores fell below 80% reliability; these were Limit Setting and Problem Identification. prot oc ols,
Upon inspection of individual
th es e ap pe ar ed to be th e re su lt of diff er ing
interp retat ions of how to score the phrase, trouble."
Reli abi lit y for the Clinic al
"get in
scales for both
groups ranged from 91 to 100%, with the exception of Anxiety for the referred group, which was 82%.
30 Chapter III
Results In order to determine whether there were significant differences between referred and nonreferred groups of chil dre n on the RATC scales,
subjects'
T scores for each of
the scales were subjected to univariate analysis with repeated measures. RAT C scales,
There was a signi fica nt main
F (1,11) = 26.35, p
< .01.
effect for
The ma in eff ect
for groups was nonsignificant, and thus an overall diffe rence betwe en groups was not supported.
However, there
wa s a si gn if ic an t int er ac ti on bet wee n gr oups and RA TC scales,
F (1,11) = 6.22, p
< .01.
Re so lu ti on 1 wa s hig her
for the referred group, while Resolution 2 was higher for the Nonreferred group, as indicated by Tukey's WSD test for diffe rence s betwe en means.
Table
1 prese nts
the mean
and
standard deviatio n of RATC scale scores for both groups.
A
summary of tests of pairwise contrasts of mean scale scores is presented in Table 2, again using Tukey's WSD criterion. To determine whether Clinical and Adaptive scores differed between groups, mean T scores were calculated for Ada pt ive and Cl in ic al
scales and an al yz ed by a 2 (Group)
x 2
(Scales) ANOVA, with scales as the withinsubjects variable. Wh il e the dif fer en ce bet we en re fe rre d and no nr ef er re d gr ou ps wa s no t sign ific an t,
ther e was a ma in
ef fe ct for the two
Table 1 Me an and St an da rd De vi at io n of Scale Sc or es Non ref er red Gr ou ps
for Ref err ed and
Group
Referred
Scales
No nr efe rr ed
Mean
SD
Reliance on Others
46.4
12.0
50.9
10.2
Support Other
38.2
11.3
37.2
9.3
Support Child
50.1
11.5
49.1
9.4
Limit Setting
54.2
12.3
52.1
11.6
Problem Identification
46.6
13.0
52.2
11.5
54.2
11.8
44.8
7.7
38.7
11.7
56.6
14.9
Anx ie ty
52.3
10.4
58.6
12.1
Aggre ssi on
55.8
10.1
52.1
12.1
Depression
63.7
13.0
65.5
10.5
Rejection
59.8
9.2
59. 0
11.8
Unresolved
67.3
14.4
59.8
13.6
Resolution 1 Resolution 2
•k
J2
<. 05 .
k
* k
k k
£
<. 01.
Me an
SD
32
Table 2 Differences in Mean Scale Scores using Pairwise Contrasts (Tukey’s W5S Criterion) Treatment
S0
R2
Rel
PI
Ri
SC
Li®
flss
flnx
Rej
Unr
Dep
S0
3.32
10.35
11.63
11.76
11.30
15.44*
16.26**
17.73**
21.68**
25.80**
26.85**
R2
1.03
. 1.77
1.84
1.38
5.52
6.34
7.81
11.76
15.88**
15.33**
0.74
0.81
0.35
4.43
5.31
6.78
10.73
14.85*
15.30**
0.07
0.21
3.75
4.57
6.04
3.33
14.11*
15.16*
0.14
3.68
4.50
5.37
3.32
14.04*
15.03*
3.54
4.36
5.83
3.73
13.31*
14.35*
0.82
2.23
6.24
10.36
11.41
1.47
5.42
3.54
10.53
3,35
8.07
3.12
4.12
5.17
Rel PI Ri s-c
Lim Ago Pnx Rej iJnr
-
Dep
(.05.
**p
(.01.
1.05
33 scales,
F
(1,
56) = 105.44,
than Adaptive scores.
£
< .01, wi th Cl in ic al
high er
The interaction was nonsignificant.
Table 3 presents the mean scores and standard deviations for Ad ap tiv e and Cli ni ca l
scales by group.
Th es e re su lt s are
contrary to the expectation that referred children would have higher mean Clinical scores. To examine the proportion of children who resolved prob le ms,
a reso lu ti on ra ti o wa s de ri ve d for ea ch su bj ec t by
com bin ing the total numb er of Res olu tio n 1, 2, and 3 raw scores and dividing by the number of opportunities available for resolution.
Mean
comp are d usin g a t 2.417, £
< .05.
scores for both groups were
test
for ind epe nde nt
groups,
then t
(56)
=
The mea n res olu tio n rati o for the
referred group was 0.43, and for the nonreferred group, 0.62, indicating that children in the latter group were more likely to resolve an identified problem situation. To address the question of correlations in general, and betw ee n RA TC scales and in te ll ig en ce or ach iev eme nt pa rt ic ul ar,
Pe ar so n*s
cor re la ti on
in
co eff ic ie nt s we re fi rs t
calculated for combined subjects, then by group, across all vari ables
except Res olut ion
3.
(Resolution 3 was omitted
be cau se it oc cu rr ed on ly tw ic e in the Ref er red gr ou p an d not at all in the Nonr efe rred group.)
In general, mean adaptive
scores were positively correlated with the Adaptive scales and mean
clinical
scores with
the Clinical
scales.
However,
Table 3 Mea n Scor es and St an da rd De vi at io ns Scales by Group
for Ada pt iv e and Cl in ic al
Scales
Ad ap tiv e
Group
Referred Nonr efe rr ed
Scale
Me an
Cl in ic al
SD
Me an
SD
Gr oup Me an
46.93
7.92
59.76
5.90
53.34
48.93
5.57
59.07
6.08
54.00
47.93
6.58
59.41
5.95
35 contrary to expectation, the Anxiety and Rejection scales demonstrated a positive correlation with mean adaptive scores
(as did Agg res sion for the referred group).
Resolution ratios, in general, correlated positively with the Ada ptive
scales.
Table 4 includ es corre lation
coefficients for selected subject variables across groups. (Correlations between resolution ratio and Resolution 1, Resolution 2, and Unresolved were omitted since those scale scores were used to compute the resolution ratio.) Full scale WI SC R scores
(IQ) for the ref err ed group,
as well as general knowledge scores from the Woo dcock Joh nson, had a po si tiv e co rr el ati on wi th Pro ble m Identification (both £ s < .01), thus supporting Roberts' observations.
The Res oluti on
cor rel ate d posi tiv ely with IQ
2 and Anx iet y scales also (both £ s < .05).
Correlations for selected subject variables for the referred group are presented in Table 5. For the non referred group all correlations between total California Achievement Test (CAT) scores and RATC scales were nonsignificant.
Aggres sion was posit ively
correlated with Clinical but not Adaptive scores for the nonreferred group.
Correlations
for the nonreferred
group
are shown in Table 6. Finally, Pearson's chi square tests were used to determine whether groups differed in the number of children
Table 4 Pearson's Correlation Coefficients for RATC Scales and Selected Subject Variabl es (Whole Group)
Ad apt ive S0
Rel
Me an Ada pti ve Score
.70
Me an Cl in ic al Score
.10
Resolution Ratio
.48
**
.71
SC
**
.03 ic ic
.53
.43
Lim
**
.09 **
.35
Scal es
.50
PI
**
.10 **
.25
.61
R —1
*
.24* *
.28
.23
R2
k
.09
*
.69
* *
.01 a
a
Clinical Scales Anx
Me an Ad ap tive Score
.42
Mea n Cl in ic al Score
.47
Resolution Ratio
.27
Ag g
** **
ic
.30 .59 .03
** **
Dep
Rej
.06
.32
.47 .02
**
.62
Unr
** **
.17
Resolution/Unresolved scores were used to calculate the resolution ratio. £
<.05 .
£
<.01.
.74 .39 a
** **
37
Table 5
Correlations Between RATC Scales and Selected Subect Variables for Referred Group
Ada pti ve Scal es S0
Rel
Me an Ad ap ti ve Score
.81
Me an Cl in ic al Score
.13
Full Scale IQ
.26
Resolution Ratio
.50
*
*
.69
* *
.07
*
.53
.49
•
.06 *
Lim
SC
*
o o
.42
.68
*
*
.17
.23 * *
*
PI
.38
.10 *
.46
.62
.52 * *
R2
Rl
*
*
.49
*
*
*
*
.00
.67 1
•
*
O
*
*
.22
.39
b
.18
b
Clinical Scales An x
Me an Ada pti ve Score
.38
Mea n Cl in ic al Score
.57
Full Scale IQa
.37
Resolution Ratio
.17
Ag g
* ** *
.49 .49
** **
Dep
Rej
.19
.32
.48
**
.61
.25
.05
.17
.15
.02
.10
Unr
* **
an = 25 Resolution/Unresolved scores were used to calculate the resolution ratio. £
<.05.
p
<.01.
.73 .45 .24 b
** **
*
38
Table 6
Correlations Between RATC Scales and Selected Subject Variables for Nonreferred Group
Ada pti ve Scales Rel
Me an Ad ap ti ve Score
50
Mea n Cl in ic al Score
10
Resolution Ratio
38
S0
**
.78
SC
**
.00 *
.63
.35
Lim
*
.21 * *
.33
*
31
*
PI
.57
03
.13
08
.28
R—1
R—2
.01
.87
. 32
.12
* *
* *
Clinical Scales An x
Mea n Ad ap tiv e Score
.44
Me an Cl in ic al Score
.46
Resolution Ratio
.23
** **
Ag g
Dep
Rej
.18
.18
.38
.66 .02
**
0 * * 0
.16
.64
Unr
* **
.28
Resolution/Unresolved scores were used to calculate the resolution ratio. * ** £ <.05. £ <.01.
i ** n
l •
.32 a
*
39 wh o pr ov ide d Aty pi cal Re spo ns es their stories. cases.
or Maladapti ve Ou tc ome s for
The results were nonsignificant in both
Mean raw scores for the refer red group were as
follows: Atypical 0.83; Maladaptive 1.17; and Refusal 0.48.
These scores we re acc oun ted
for by 41%
(12),
55%
(16), and 24% (7) of the referred subjects, respectively. Mea n ra w scores for the non ref er red At ypic al 0.52; Mal ada ptiv e 0.45;
gr ou p wer e as foll ow s: and Re fu sa l 0.03.
Thes e scores wer e accou nte d for by 34% 3% (1) of the non ref err ed subjects,
(10),
34% (10), and
res pecti vely.
Thus,
while the re su lt s we re stati st ic al ly in si gn if ic an t,
the
overall pattern of scores on the Clinical Indicators was similar to that of the original research population.
40 Chapter IV
Discussion Discussion of the Results The basic question of the present study was whether the RATC would discriminate between two groups of students: those who displayed problems of learning and/or behavior to the extent that they were referred for a psychological evaluation, and children who had not been referred and therefore were generally considered to be "normal."
In
quantitative terms, the results were only partially in the expected direction.
The principal differe nce which emerged
wa s in the ab ili ty of no rm al ch il dre n to re so lv e pro ble m situations
in their
stories.
wi th Mc Ar thu r and Roberts'
This find ing is consi sten t
(1982)
obser va ti on that
cl in ic
children tended to be less mature cognitively and thus frequently resolved problems in a quick and easy manner (i.e., Resolution 1). In neither the original study nor in the present one did the clinical scales of Aggression, Anxiety, or Depression differ significantly between
groups.
The authors
of the test suggested that while individual protocols were sensitive to differences in affect (including feelings such as anxiety and depression), such differences tended to "wash out" when averaged into the data of a relatively
41 heterogeneous
group.
One woul d expect Roberts' mark ed di ff er en ce s,
rese arch sample to disp lay
si nce th e st an dar di zat ion gr ou p wa s
comprised of children nominated on the basis of being wellad justed
(accor ding to se ve ra l
sp ec if ie d cr iter ia ),
and
the clinic group, children referred to Child Guidance Clinics.
It appears
that a rando mly drawn publi c school
sample does not differ dramatically from classmates who have be en re fer re d to the scho ol ps yc ho lo gi st .
A co nfo und ing
factor in the present study was that the research sample was a heterogeneous group of referred students; some children did not qualify for special services (although they evidenced problems which were serious enough to indicate the need for an evaluation).
Furthermore,
be ha viorally im pa ir ed ch il dr en
while all
(by defi ni ti on )
and many
learning disabled children evidence some emotional problems, certainly there are students who appear welladjusted in spite of academ ic difficulties.
Finally,
the suburba n
location of the school district involved in this study included a relatively restricted range of subjects in terms of ethni city and socio econ omic
level.
This combi nati on of
factors may well have minimized the differences between groups. One shortcoming of this study was the lack of secon dary
students in the nonref erred
sample.
Thus
it was
42 not possible to examine Resolution 3 scores between groups. In the referred group, however, 1215 year olds had a total of only two Resolution 3 scores, with a preponderance of Res olut ion
1 responses.
When the pro port ion of resolve d
stories was compared (using a resolution ratio), the nonr eferr ed
students did score highe r as a group.
This
finding seems consistent with McCona ughy and Ritter's
(1986)
research, which indicated that learning disabled boys demonstrated more problems in social competence and behavior than normal boys the same age. In regard to the positive correlations between some of the clinical
scales
(Anxiety and Rejecti on for both groups,
as well as Aggression for the referred group), some conjec ture may be in order.
Roberts points out that the
ability to express a certain amount of aggression is healthy, and several of the RATC cards specifically "pull" for aggres sion. beating)
It is the aty pica l exp res sio n
(e.g.,
or de ni al of ag gr ess ive fe el in gs whi ch is
clinically significant in such instances. An xiet y is poten tia lly ada pti ve in moder ate am ou nt s also, in terms of optimal arousal theories (i.e., either too mu ch or too li tt le is debilit ating .)
Similarly,
in the
Koppitz scoring method for the Bender VisualMotor Gestalt test, constriction is considered a sign of anxiety, but has be en shown to co rr el at e posi ti ve ly wi th school ac hi ev em en t.
43 There may be an unintentional confounding of the Anxiety scale, too, in that making an apology (which is scored on the Anxiety scale) may be quite appropriate in some situations. It is not clear from the available data whether the pos iti ve co rr el at ion be tw ee n IQ and Pro ble m Id en ti fi ca tio n (PI) is due to the ability of brighter children to express themselves verbally, or to the cognitive ability to identify a problem.
McAr thur
and Roberts
state that childr en with
high PI and low Resolution scores are good candidates for therapy because they at least are able to articulate pr ob lems .
It wi ll no t su rp ri se
scho ol psych olo gist s to
learn that the general knowledge subtest of the Woodc ock John son co rr el at ed posi tively wi th PI, and thus wi th IQ as well?
a patt er n of low scor es
the achiev ement areas mat h)
in on e or more of
(reading and wri tten
language, or
and a relativ ely hi gh sc or e on ge ne ra l kn owl edg e is
virtually a hallmark of learning disabled youngsters. Ac ade mic ac hi ev eme nt
scores,
on th e ot he r hand,
ap pea re d to
have no consistent relationship to the RATC. The Clinical Indicators are intended to act as warning signals of emotion al disturbance.
Again,
the het erog ene ity
of the samples in this study may have obscured real differences between groups. group,
for example,
Two students in the nonreferred
acco unt ed for 11 respo nses.
Thus the
44 nonreferred sample included children who showed evidence of emotional disturbance although they did not exhibit unusual pr ob le ms
in school,
just as some le ar ni ng dis ab led
st ud en ts
demonstrated problems in academic but not emotional areas. Clinical Validity Having examined the RATC from the viewpoint of statistical analysis, however, a crucial question remains regar ding where
its val idi ty as a clinical
instrument.
In cases
sp ec if ic be ha vi ora l pro ble ms we re al rea dy evident,
pa re nt s we re
the
so me ti mes as ke d to re sp on d to the Pe rson ality
Inventory for Children (PIC; revised format profile form) (Wirt, Lachar,
Klinedin st,
& Seat,
we re av ai la bl e for ch il dr en in this
1982).
Five PIC pro fil es
study,
and while not
subjected to statistical analysis, a comparison of PIC and RATC profiles revealed that the same areas of concern were identified in nearly a onetoone correspondence in four of the five cases.
(In the fifth instance,
th e fat her
served
as respondent for the PIC; it was the examiner’s impression that the daytoday care of the children was left to the stepmother in that family and he was unaware of his child's emotio nal
st at e. )
This
is a form of conve rgent
validity,
since two very different types of information parental observations of behavior and projective material provided by the child pointed to the same conclusions. Several examples may help to demonstrate the usefulness
45 of the RATC as a clinical instrument.
McA rth ur and Roberts
note that very low scores on the Clinical scales may reflect denial
or avoidance.
This was clea rly illust rated
in the
case of a 15 year old boy who had at one time been in an engineered classroom
for the behavior ally
several years with a foster family,
impaired.
After
he had made good
pr og res s and was gen eral ly do in g we ll
in ju nio r hi gh
school.
He had been suspended on four occasions the previous year, however,
for outbu rsts
of aggres sive
behavior.
On Card 13,
whi ch de pi ct s a bo y wi th a ch ai r rai se d ab ov e his he ad in anger,
he responded as follows:
"He's carr ying a chair home
from a garage sale and he's just real happy, and since he's in a good mood, he's kinda hyper and lifting it over his head."
Similar nonaggress ive responses were given for cards
whic h pull for ag gr es si on .
This case
su pp or ts the pl ea
for
nonambiguous stimuli (Epstein, 1966)? when the expected respo nse is known,
clearer conc lusio ns may be drawn from
respons es whic h are a departu re
from the norm.
(The
perc ent age of cl as se s of re sp on se s for ea ch ag e gr ou p are listed in the RATC manual, which is helpful when determining how typical a particular response might be.) It was noted that many responses were autobiographical in nature. for Car d 12:
For example, "This girl
an 11 year [woman]
they had a big figh t and she
old girl told this
story
is lea vin g this guy beca use
[the child] wa s watc hing.
She
46 wa s su pp osed to be in her ro om do ing homework,
and th ey had
a big fight over whose house it is and who's going to pay the rent,
and she
(the lady]
just de ci de d to leave ."
The
mo the r had ta lk ed with us pr ior to th e ev alu at ion and related
similar information;
month later.
in fact,
she did leave one
The ch il d had be en re fe rr ed bec au se of fa il ur e
to complete assignments. A 13 ye ar old boy, family,
wh os e mo ther had rec ent ly left the
told this st ory to Card
2:
"This wom an is happ y to
see her son because she has been divorced and her son was living with her (sic) father and she hasn't talked to him on the phon e or even got a letter wr it te n to him.
(End?) Don' t
know the mom pro bab ly stays with hi m a little while,
then
goes back to where she's living because they're divorced. Ma ybe her son goes to live wi th her, she'll probably keep him anyways."
for a li tt le while This story reflects both
the reality of the separation and the boy's longing for his mo th er.
The fa th er told us that the mot her wa nt ed the
youngest child to live with her while the three older ones remained with him. A 10 ye ar old boy wh o came fr om an ab us iv e fa mi ly and who ex hi bi te d nu me ro us
at yp ic al be ha vi or s in the cla ssr oo m
demonstrated problems with reality testing in his response to Card 15:
"He's watchi ng his mom take a bath the boy
and then someone rang the doorbell and he shut the door and
47 look ed and no one was there. him to leave. holdin'
It was
some soap,
It was
just his mo m gettin'
just a bal loo n or a pic tur e of her
or a wate rpr oof
robot."
Anot her
8 year
old who displayed a great deal of acting out behavior at home and at school responded in an atypical fashion to many cards.
Here is her res pon se to Card 7: "Once upon a time
ther e was
a girl who
found
a boyfr iend.
(These guys
are 19
or 20.)
Then she wok e up and saw her boyfriend c oming in.
He said,
'Wanna get mar rie d?'
She said,
'Well, O.K.'
Then
they got into bed together and they kissed, and then her mom came in, and she had to hide Johnnie under the covers. he started to kiss her legs. bedroo m.
Then her mo m got out of the
Then he go t out of the covers,
married un de r the bed,
Then
and th ey got
wi th all the mice ."
This is a ch il d
for whom the mother's PIC reflected spikes on numerous scales. A pa ir of stor ie s wi ll be pre se nte d to il lu st ra te differing
abilities
to resolve problems.
Although both
students,were being reevaluated for learning disabilities, the girl comes from a non su pp or ti ve , abusive fa mily while the boy has a very strong,
close knit family.
He also has a
higher IQ (113, compared to 88 for the girl), which suggests the cogni tive dimen sion invol ved in prob lem solving. girl, ag ed 14, respo nde d as foll ows to Card 3:
The
"She's tired
and she don't want to do her homework and it's due tomorrow,
48 and she's fall ing asle ep."
The boy,
the same card in this manner:
aged 12, res pon ded to
"This kid
just got home from
school, and he's mad at the teacher 'cause she gave him so mu ch home work,
and he's an gr y and th ro ws
his pe nc il on the
ground, and the more he thinks about it, the more he gets mad.
An d he goes and talk s to his mom ,
pati ence ,
yo u' ll get it all done,"
and she says,
"Have
and he goes ba ck and
finishes all his homework, and then he goes to school the next
day,
and he*g ets
l's on all of his homewo rk.
That' s a
good way to end a story." It was clear,
in revi ewing the protoco ls
from this
study, that all children had problems. As the statistical results demonstrated, the difference between groups frequently lay in the way the proble m was resolved.
A
nonreferred fifth grade boy of Filipino descent told this story for Card 6:
"Well, these two kids mi gh t not let this
kid play because he's a different color than them, and they mi ght be pre jud ic ed of
'im.
Before,
he might've be en ne w to
the school, and he might've felt strange around other color people ,
and th ey maybe ma ke
diff eren t color.
fun of hi m bec aus e he's a
He can mayb e talk to his parents or the
pr in ci pal ab ou t it, and th en he can be co me
fr ie ndl ie r and
maybe th es e kids can le ar n to like hi m cuz if yo u' re a diffe rent
color,
that
doesn't mean you'r e
had rec ently moved f rom the east coast,
so bad."
and no doubt
This boy
49 included more than a little autobiographical information in his story.
A referred fou rth grade boy told a story for the
same picture as follows:
"The boy's goin g to the park, and
these two kids stopped him and said,
'Give me your m o n e y , '
cuz there 's a candy store at the park.
So these kids took
his money and they got something and he didn't, and they be at hi m up, an d he got a bi gg er ki d and be at him the big kid's not in the picture.
up bu t
Then the kid beat him
u p ." Conclusion It seems clear that a projective test such as the RATC contributes valuable information to a psychological evaluation.
In additi on to the studies of reli abi lity and
validity referred to earlier, the present study tends to support convergent validity based upon the PIC and other less formal measures such as interview and sentence comple tion techniques.
Togethe r these components of an
evaluation lead to a broader, more integrated view of how a given
child is curre ntly functioning.
While the resul ts of
this study indicate that "normal" kids, at least in one suburban school district, tend to score less differently from their counterparts with learning and/or behavior pr obl ems
th an mi ght be expe cted,
the RAT C still may be a
valuable tool in discovering what factors enter into a par tic ula r chil d' s situation,
and he lp
id en ti fy his /h er way s
50 of perceiving and responding to problems. Using more clearly defined diagnostic subgroups would be he lp fu l in fu tu re re se ar ch of this kind, larger and more dive rse samples of
"average"
is clear from the present study, however,
to ge th er wit h students.
It
that children with
pr ob le ms ne ed to le ar n problems olvi ng skills.
Sp ec if ic
teaching in that area appears to be as essential as remediation in content areas, and certainly its impact will extend well beyond the classroom.
References Adcock ,
C. J.
tests
(1970).
and
In O. K. Bu ro s
reviews
(PP
(Ed.),
12291230).
Pers onal ity
Highland
Park,
N.J.:
Gryphon Press. An as ta si , York: An drew,
A.
(1976).
Ps yc hol ogi ca l te st in g
(4th
Ne w
Macmillan.
G., Walton,
(1953).
R. E., Ha rtwe ll ,
S. W . , & Hutt,
The Mic higa n Pictu re Te s t . Chicago:
M. L.
Science
Research Associates. Beliak,
L.
(1954).
The Thema tic App erc ept ion
Test and the
Ch il dr en 1s Appe rcep tio n Test in clinical u s e . New York: Grune & Stratton. Beliak,
L.
(1968).
Discussion:
The Child ren's
Apperc ept ion
Test: Its use in developmental assessments of normal children.
Journal of Proj ectiv e Techniq ues and
Pers onal ity As se ss me nt , 32, 425427. Belia k,
L . , & Ade lma n,
C.
Ap per ception Test. (Eds.),
(1960).
The Chi ldr en 's
In A. I. Rab in and M. R. Ha wor th
Proje ctive tech niques wit h child ren
(pp.
350363). New York: Grune & Stratton. Beliak,
L., & Beliak,
Tes t Beli ak,
S. S.
(C A T ). Ne w York:
L . , & Beliak,
S.
S.
(1949). CPS,
Ch il dr e n's App er ce pt io n
Inc.
(1950). An
the Childr en's App erc ept ion Test
intr od uc to ry no te on (CAT).
Projec tive Tec hn iq ue s, 14, 173180.
Journal
of
52 Beliak ,
L . , & Hurvi ch,
Apperce ptio n Te st
M.
S.
(1965).
C h i l d r e n 's
(Hu ma n F i g u r e s ): Man ua l for the
CA T H. Los Angeles: We ster n Psychol ogic al Services. Blum, G. S.
(19 50).
The Bl ac ky Pi c t u r e s : A te ch ni qu e for
the expl orat ion of per son alit y dy na mi cs . New Y o r k : The Psychological Corporation. Cull inan ,
D . , Epst ein,
M. H . , & Mc Lin den ,
D.
(1986).
Sta tus
and change in state administrative definitions of beh avi or di sorder .
Sc hool Psyc hology Review,
15,
383392. Epstein,
S.
(1966).
Some theor etica l consi dera tions on the
nature of ambiguity and the use of stimulus dimensions in projective techniques. Psychology, Frank, L. K.
183192.
(1939). P roj ect ive me th ods fo r the study of
pe rs on al it y. Haw ort h,
30,
Journal of Consulting
M. R.
Jou rn al of Ps yc ho lo gy ,
(19 66 ).
8,
38940 9.
The C A T ; Fa ct s abo ut f a n t a s y . Ne w
York: Grune & Stratton. Hersh,
J.
(1978).
A concu rrent vali dat ion
of the Roberts
Apperc epti on Te st as a me asure of ther ap eut ic pr og re ss of ch il dr en . Unpub lish ed docto ral
dissertation,
Cali forni a School of Prof essi onal
Psychology.
Diss erta tion Abst ract s In te rn at io na l, 1979, 40324033B)
(From 39,
Hertz, M. R.
(1970). P roj ect ive tec hni que s in crisis.
Journal of Projective Techniques and Personality A ss e ssmen t, 3 4, 4494 67. Holt,
R. R.
(1950).
Review:
C hil dre n's
Appe rc ept ion
Journal of Proj ectiv e Tec hn iq ue s, 14, Jas tak ,
J. F . , & Jas tak,
S. R.
(1976).
198200.
The Wi de Ra ng e
Ac hie ve men t T e s t : Man ua l of in st ru ct io ns Wilm in gt on , Klopfer,
W. G.
meth od s.
Test.
(r e v i s e d ).
DE: Gu id anc e Asso ci at es of Delawa re.
(1968).
The me ta mo rph osi s of pro jec tiv e
Journ al of Pro je cti ve Te ch ni qu es and
Pers onal ity As se ss me nt , 32, 402404. Klop fer,
B.
& Dav idso n,
H. H.
(1962).
The Ro rs ch ac h
te ch ni qu e: An introd ucto ry ma n ua l. New York: Har court Brace Jovanovich. Ma gn us se n,
M. G.
(1980).
In R. H. Woo dy ass ess men t Mc Ar th ur ,
D.
(pp.
(1976).
Ch il d an d ado les cen t de ve lo pm en t.
(Ed.), 116).
Enc ycl ope di a of clin ical San Franc isco:
Jos seyB ass.
A co mp ar is on of the st im ul us
in fl ue nc e
of three thematic projective techniques with children. Disse rtati on Abst ract s
Int er na ti on al, 37,
(University Micr ofilm s Intern ational, McAr th ur ,
D. S.,
& Ro be rt s,
G. E.
(1982).
09B.
No. DBJ 776307) Ro be rt s
Apper cept ion Test fo r Ch il dre n m a n u a l . Los An ge le s, We ste rn Psy cho log ica l Services.
CA
Mc Co na ug hy ,
S., & Ritter,
D.
(1986).
So ci al co mp ete nce and
be ha vi or al pr ob le ms of le ar ni ng di sa ble d bo ys aged 611. McKi nn ey ,
Jo urnal of Lea rni ng Di sa bi li ti es , 19, J . , & Feagans,
L.
(1983).
be ha vio r of lea rn in g di sa bl ed
3945.
Ad ap ti ve cl as sr oom students.
Jo ur na l of
Lear ning Di sa bi li ti es , 16, 360367. Mc Ki nney ,
J. D . , & Speece,
D. L.
(1983).
Cla ss roo m be ha vi or
and the academic progress of learning disabled students. 4,
Journal of Applie d Deve lopm ental
Psychology,
149161.
Mo ri ar it y,
A. E.
(1968).
No rm al pr es cho ol er s'
re ac ti on s to
the CAT: Some implications for further development. Journal of Projective Techniques and Personality A s s e s s m e n t , 32, 4134 19. Muha,
T. M.
(1977).
A va lida tio n st ud y of the R o b e r t s '
Appercep tio n Te st as a measure of ps ych olo gi cal dysfunction in fa mil ies . Unpublished doctoral dissertation, California School of Professional Psychology.
(From
Dissertation Abstracts
In te rn at io na l, 1977, Murray ,
H. A.
(19 43).
38,
5034B)
Th em at ic Appe rception Test m a n u a l .
Cambridge, M A . : Harvard University.
Murs te in ,
B. I.
(1968).
pe rson al ity,
Ef fe ct of stimulus,
and sc or in g sy st em on the manifestation of
host ilit y on the TAT. Clinical Mu rs te in ,
Psychology,
B. I.
ba ck gr ou nd ,
(1970).
Journal 32,
of Cons ult ing and
355365.
Review.
In O. K. Bu ros
Perso nalit y tests and reviews
(pp.
(Ed.),
1229f.).
Highla nd
Park, N. J . : Gr yp ho n Press. Ne ur inge r,
C.
I. Rabin
(1968).
A va riet y of th em at ic me thod s.
(Ed.),
Proje ctive techn iques
asse ssm ent Newv il le,
L.,
(pp.
& Hamm,
222261). N. H.
N ew York:
(1985).
In A.
in per son alit y
Springer.
Co nt en t validity of the
WRA T and grad e level sch ie ve me nt of pr im ary gr ad e students. Pihl,
Jou rnal of School Psycho logy,
R. O . , & McL arno n,
L.
D.
children as adolescents.
(1984).
23,
9193.
Le ar ni ng di sa bl ed
Journal of Lear ning
Di sa bi li ti es , 17, 96100. Rabin,
A.
I.
(1960).
Pr oje cti ve met hod s and pr oje cti on
in
children. In A. I. Rabin & M. R. Haworth (Eds.), Projective techniques with children (pp. 211). New York: Grune & Stratton. Rabin,
A.
I.
(1968).
intro duction .
Pro jec tiv e metho ds:
In A. I. Rab in
(Ed.),
An his tor ica l Pr oje cti ve
techniques in personality assessment (pp. 317). New York:
Springer.
Roberts,
G. E.
(1958).
The eff ect of stim ulus va ria tio n on
the projection of ch il dr en . dissertation,
Purdue University.
A b st r a c t s , 1958, Sanford,
R. N.
(1943).
environment.
Unpublish ed doctoral Dissert ation
19, 365366) Stud ies of pe rs on al it y and the
Monograp h of the Society for Researc h in
Ch il d D eve lo pm e nt , 8, 2583 01 She lto n,
(From
(No. 1, Ser. No.
T. L . , An as t op ou l os , A. D . , & Linden,
43).
J. D.
(1985).
An att ribu tio n tr ai nin g prog ram wi th le ar ni ng di sa bl ed children.
Journal of Lea rni ng Di sa bi li ti es , 18,
261265. Shneidm an,
E. S.
(19 49).
The Ma ke A P ic tu re S to ry T e s t . New
York: Psychological Corporation. Symonds,
P. M.
(1948).
Manu al
for S ym o nd s ' Pict ure
Sto ry
Te st . New York: Co lumb ia University. Wirt,
R. D.
(1970).
Review:
CAT.
In O. K. Bu ro s
(Ed.),
Personality tests and reviews (p. 1230). Highland Park, N. J . : Gryp ho n Press. Wirt,
R.D.,
La char,
D., Kl in ed in st ,
J . , & Seat,
Perso nal ity Inven tory for Chil dren
P.
(1982).
(P I C ). Los Angeles:
Wester n Ps ych olo gic al Services. Wi th er sp oo n,
R. L.
(1968).
Devel opme nt of ob je cti ve
for longitu dinal CAT data. T echn iques , 3 2., 406 412.
sc or in g
Journ al of Proje ctive
57 Zubin,
J . , Eron,
L. D . , & Shumer,
F.
(1965).
An
experimental approach to projecti ve tec hn iq ue s. New York: Wiley & Sons.
58 Appendix A
Car d 1B Family Confrontation
Card 2G M a t e r n al S u p p o r t
Card 3B School Attitude
Card 4 Child Support/Aggression
maw Card 5G Parental Allection
Card 9 Physical Aggression
Card 6B P e e r / R a c i a l I n t e r ac t i o n
Card 7G Dependency/Anxiety
Card 11 Fear
Card 10B Sibling Rivalry
Card 8 F a m i ly C o n f e r e n c e
Card 12G Parental Conflict/Depression
m 'V ~ v Card 13B Aggression Release
Card 14G Maternal Limit Setting
The RA TC
Card 15 N ud l ty / Se xu al lt y
S t i m u l us C a r ds
'?■
Card 15B Parental Support
59 General Description of Stimulus Cards
There are 27 stimulus cards, of which only 16 are administered to any given child. The 11 cards designated B/G indicate those with parallel male and female versions.
The
typical themes elicited by each card are described below (as ad ap ted
from Mc Ar th ur
& Rober ts'
Ma nu al
, pp.
2 & 4).
1B/G (Family Confrontation) usually elicits stories in which pa re nt s gi ve ad vi ce or co rr ec t a wr on gd oi ng .
A ch ild's
response to this card may help clarify the presenting problem,
es pe cia lly th e na tu re and ex te nt of pr ob le ms
in
family relationships. 2B/G (Maternal Support) usually evokes stories in which something traumatic or negative has occurred and the child is reach ing out to the moth er for help.
Response s
ma y re ve al the na tu re of the chil d' s depend ency ne ed s in relation to a maternal figure, and how she responds. 3B/G (School Attitude) usually reflects a child's attitudes toward
school.
out for help,
It may reveal whet her
a child reaches
from wh om supp ort is availa ble,
and the
level of the child's academic aspirations and successes. 4
(Child Support /Aggressio n)
strong ly suggests
anteceden t
themes of aggression, accident, or illness, and usually reveals the nature of support given by the standing girl
60 and others who may be called on for help. 5B/G (Parental Affection) focuses on the child's observation of intimacy between male and female adults, especially parents, warm th,
and of te n re ve al s fe el in gs
of re je ction,
or jealousy.
6B/G (Peer/Racial Interaction) reflects the child's interactions with peers including friendship, rejection, or rivalry.
Racial attitudes may be expressed because
one of the peers is black. 7B/G
(Dependency/Anxiety)
frequently
evokes
stories
about
anxietyproducing situations such as waking up from a bad dre am,
fear of the dark,
or fe el in g ill.
It may
pr ovi de clues as to whether th e ch ild tr ies to cope wi th anxiety alone or seeks the help of others. 8 (Family Confer ence)
elici ts a wid e va ri ety of stories,
including punishment for the children for something they did wrong, or planning something positive to do together. mo th ers
Children being raised by singleparent
fr equ en tl y see the ma le ad ul t fi gu re as an
auth orit y figure
(e.g., doctor, principal, m inister).
9 (Physical Aggr essio n Toward Peer) p rovides an oppo rtu nity for children to express their feelings about aggression, their need for help, and the coping skills they use for resol ving conflicts. significant.
Denial of aggre ssio n is cli nical ly
10B/G (Sibling Rivalry) tends to elicit a child's feelings about having a new sibling including jealousy, curiosity, a wish to care for the baby, and concern about the moth er's
continu ed
love and availability.
The
response of the mother may yield clues about the child's pe rc ept io n of the mo th er' s pa re nt in g. 11 (Fear) often elicits stories about the girl reacting to an external danger such as an animal or another person, or some trau mati c event.
Childr en may reveal their
ability to cope with fear by themselves or their tendency to call on others for help in solving a problem . 12B/G (Parental Conflict/Depression) generally elicits themes of paren tal
conflict or depression.
Childr en may
interpret the card as father's response to mother's illness, parents arguing, or parents upset about an external
event or someth ing the child did wrong.
The
stories usually detail the father's role in providing support and the child's reaction to parental upset. 13B/G (Aggression Release) usually evokes details of what happened to make the person angry and how he/she will handle the aggre ssive
feelings.
Denial may
suggest
discomfort with aggression or a need to avoid expressing aggression. 14B/G (Maternal Limit Setting) generally elicits a child's
62 response of wrongdoing and subsequent parental pu ni sh ment .
It may re ve al
clues ab ou t ho w li mi ts are
set in the family, and the nature and severity of the pu ni sh ments . 15 (Nudity/Sexuality) generally reveals a child's emotional reacti ons
to nudit y and to sexua lity more
the narrative,
the boy may
generally.
leave, be come embarrassed,
In or
express curiosity; the girl may react with anger, embarrassment,
or seek paren tal
intervention.
The child
telling the story may show behavioral signs of embarrassment. 16B/G (Parental Support) typically reveals the nature of the relationship between the child and the father or a father figure,
and may provide
information about how the
child perceives the father's parenting.
