CHAPTER 2shodhganga.inflibnet.ac.in/bitstream/10603/36631/6/06_chapter2.pdf · Review of Literature...
Transcript of CHAPTER 2shodhganga.inflibnet.ac.in/bitstream/10603/36631/6/06_chapter2.pdf · Review of Literature...
Review of Literature
60
• Chapter-2 consist of following areas:
1. Form Appropriate Extended (XA%) and Form
Appropriate Common Areas (WDA%)
2. Conventional Form (X+%) and Unusual Form (Xu%)
3. Distorted Form (X-%)
4. Popular Responses (P)
5. Reliability, Validity and Norms in relation to
Cognitive Mediations
5.A. Reliability
5.B. Validity
5.C. Norms
Review of Literature
61
Present study is conducted to examine the Cognitive Mediation of
Participants with Mania with Psychotic Symptoms, Severe Depression
with Psychotic Symptoms, Undifferentiated Schizophrenia and Normal
Healthy subjects. There are few studies have done in this concern.
Cognitive Mediations consists of Form Appropriate Extended (XA%),
Form Appropriate Common Areas (WDA%), Conventional Form (X+%),
Unusual Form (Xu%), Distorted Form (X-%) and Popular Response (P).
Following Studies have been done in this variables.
1. Form Appropriate Extended (XA%) and Form Appropriate
Common Areas (WDA%):
XA% is a calculation of responses that are viewed as positive and
work conjunction with WDA%. XA% represents an individual’s
responses to the contours of the inkblot. These responses are viewed as
conventional because the individual only used the shape of the inkblot
to describe what they see.
The calculation for XA% is anticipated to be large and similar to
WDA% but WDA% will usually have a superior percentage. However,
there are situations in which the converse may occur. Interpretations of
the pairing will rely on the calculation of each variable and the extent of
the difference between the pair. Exner (2003) stated that if XA% is
between .78 to .90 and the value for WDA% is equal to or greater than
XA%, this is indicative that mediation is usually appropriate for the
situation, or the subject possesses intact reality testing. When XA% is
less than .70 and WDA% is less than .85, it suggests that a tendency
toward mediational impairment is somewhat pervasive. When the XA%
Review of Literature
62
is between .70 and .77 and the WDA% is between .75 and .79 it
indicates a moderate level of meditational dysfunction. When the XA%
falls below .70 and the WDA% is between .75 and .79, it indicates a
more substantial level of dysfunction. When the XA% is less than .70
and the WDA% is less than .75, it reflects a significant meditational
impairment. Two elements are important to understanding how
pervasive the impairment mediation may be. The first is the value for
the WDA%. When the WDA% falls between .65 and .74, the dysfunction
is serious and reality testing will be noticeably affected. When the
WDA% is below .65, the dysfunction is severe and reality testing will be
markedly impaired. The second element of concern is the difference
between the XA% and the WDA%. It often provides information about
the extent to which the impairment to reality testing will impact on
everyday functioning. When the difference between the two values is .10
or more, it suggests that the dysfunction will be more noticeable in
circumstances where cues to mediation are less obvious. When the
difference between the two variables is less than .10, it can be assumed
that the impairment is global, that is, the dysfunction tends to occur
regardless of how obvious distal cues may be.
After a thorough research review minimal research was found for
Form Appropriate—Extended (XA%). Due to the paucity of research
available it is believed by the researcher to be important to investigate
this variable further. XA% was introduced in 2003 when Exner elected
to adapt the Cognitive Mediation cluster in order to enhance the
understanding of client mediation. This calculation is achieved by
Review of Literature
63
dividing the sum of responses that have a Form Quality coding of F+,
Fo, or Fu (“good form fit” responses) by the number of responses (R) in
the protocol (Exner, 2003). F+ consists of answers that would normally
be scored “ordinary” but the response has been improved without
violating the feature of the response. Fo are responses in which the
respondent has communicated form features which are easily
identifiable. These answers are consistently found in 2% of subjects
from the data pool when using W and D areas, or by 50 people in the
pool who responded to Dd areas. Fu are answers that tend to be
uncommon but are seen quickly by an observer. The basic contours
have not been violated and are appropriate. XA% is a calculation of
responses that are viewed as positive and work in conjunction with
WDA%. XA% represents an individual’s responses to the contours of the
inkblot. These responses are viewed as conventional because the
individual only used the shape of the inkblot to describe what they see.
Exner (2001) stated that Form Quality was worth investigating because
individuals under stress or experiencing pathology may or may not
maintain an accurate perception of reality. The calculation for XA% is
anticipated to be large and similar to WDA% but WDA% will usually
have a superior percentage. However, there are situations in which the
converse may occur. Interpretations of the pairing will rely on the
calculation of each variable and the extent of the difference between the
pair.
Le (2002) included XA% and WDA% in her Vietnamese sample
size of 27. She found a significantly low XA% (.65) and a WDA% of .69.
Review of Literature
64
The current norm set by Exner for XA% is .92 and WDA% is set at .94
(2001). The calculation in assessing these two reality testing variables
would lead one to assume that Le’s highlyeducated sample exhibited
tendencies toward poor reality testing. However, she believed that her
student status, her sample’s empathy for her dissertation plight, or a
cultural factor of wanting to please her may have led to an increase in
the number of overall responses and thus affected her results.
2. Conventional Form (X+%) and Unusual Form (Xu%):
The mean X+% for non patients both children and adults between
.70 to .80. Standard deviations are 0.10, while standard deviations for
various patients groups tend to be higher than non patients. The X+%
means for patients groups are lower than for nonpatients. It is .64
(SD=.14) for a heterogeneous group of 535 outpatients, .53 (SD=.12) for
279 inpatients diagnosed as having serious affective disturbance, and
.40 (SD=.15) for 200 inpatient schizophrenics.
Conventional Form (X+%) represents the proportion of ordinary
responses generated. High responses, greater than .85 reflect behavior
patterns consistent with compliance to social demands. The proportion
of answers that do no violate appropriate use of the blot contours but
do reflect less common ways of translating the stimulus field is scored
as Unusual Form (Xu%). A high responses, greater then .25, implies a
likelihood to be ‘overlly individualistic’ and percentages less than .10
suggest some mediational difficulties (Exner, 2003). X+% that falls
between .70 and .85 in combination with an Xu% less than .10 is
viewed as tendency toward mediational dysfunction (Exner,2003).
Review of Literature
65
Exner (2001) stated that Conventional Form (X+%) is the
percentage of all the responses that receive a Form Quality coding of F+
or Fo. These are appropriate responses. An X+% mean more than .85
indicates a great degree of conventionality, no matter what the other
variable values indicate. It may also suggest a fixation with social
conformity that forfeits individuality. X+% can also be a useful
measurement of obsessiveness or perfectionistic tendencies (Exner,
2003).
Conversely, X+% with a mean between .55 and .69 and an Xu%
mean of .20 or greater can indicate that the individual tends to make
more decisions that disregard social expectations than the general
public. This unique individual can display antisocial tendencies or they
may be more autonomous in their mediational processes. An X+% less
than .55 introduces the interpretive importance of considering X-%.
When X-% is greater than .20, a greater tendency of abnormal behavior
patterns in mediational dysfunction will probably be observed. Any
orientations toward autonomy and individuality should be avoided in
the interpretation. But, these calculations are not suggestive of poor
reality testing alone just that the subject does not rely on societal
expectations or demands to function. These percentages can contribute
to understanding the subjects’ presenting problem. Exner (2003) has
found X+% to be a reliable variable with high consistency. It is the only
variable within the Rorshach Comprehensive System consistently high
with nonpatient children. However, empirical studies have indicated
Review of Literature
66
that a low X+% can lead to interpretative errors with vulnerable
populations.
In Dadario’s (2002) study, she investigated five Rorshach
variables that included X+%. The participants were twelve nonverbal
learning disabled individuals (with a mean age = 14.88). Significant
differences were found between the participants and Exner's normative
sample of age-matched children and adolescents. She then compared
her group to Exner's adolescent inpatient schizophrenic sample data
and found no statistically significant differences. Based upon these
results, Dadario felt this could lead to an increase in false positives and
stated that improving conceptualizations and interventions with such a
vulnerable group should be explored further.
Locke (1999) found in her study comparing nineteen ADHD
diagnosed adults published norms that X+% was found to be
statistically significant. Conclusions were drawn that ADHD adults are
different in perception and conventionality and they also run the risk of
receiving a false interpretation because of X+% differences. However,
Smith,et. al (2002) found an increase in X+% in their study of twenty-
two, well-adapted transsexuals who had undergone sexual
reassignment surgery. Pre-surgery means for X+% was .50 and the
post-operative mean was .62. The authors attributed this spike in
Conventional Form to a decrease in X-% and Xu% after the surgical
procedure. Smith revealed in her limitations, however, that her sample
possessed parents who were supportive of their child’s surgical
procedure. It is this perceived support or results of receiving sexual
Review of Literature
67
reassignment surgery that could partially explain the inconsistency of
X+% for this sample. In identifying severe and pervasive mental
illnesses, Exner considers X+% to be a valuable consideration that
discriminates nonpatients from more seriously disturbed psychiatric
groups (Exner, 1993). The mean X-% for nonpatient children and adults
tends to be .78. While those with more serious illnesses, like
schizophrenia, are apt to display a mean of .40.
The other variable of the Cognitive Mediation cluster to be
explored is Xu%. The calculations are derived by dividing the number of
Form Quality responses coded as unusual by the number of responses
for the record (Exner, 2003). These appropriate responses tend to occur
with low frequency and can be seen quickly seen by the examiner
(Exner, 2001). The Comprehensive System’s non-clinical, normative
adult sample has a mean of .07, non-clinical sixteen year olds tend to
have a mean range of .15, for five year olds this is likely to be around
.21, and the normative schizophrenic sample has a Xu% between .17
and .21. The range of Xu% responses can either indicate a departure
from conventional behavior or a tendency to accommodate to societal
expectancies. So, for interpretive purposes the direction of Xu% in
relation to X+% provides useful information about the subject’s
mediating behavior patterns. An X+% falling between .70 and .85 with a
Xu% between .10 and .20 reflects a behavior style that is in accordance
with societal expectations. An X+% between .55 and .69 combined with
an Xu% of .20 or more reveals a person who is likely to disregard social
rules. Finally, when X+% is less than .55 and Xu% is greater than .20,
Review of Literature
68
it is likely that mediational dysfunction is occurring and reality testing
is poor. Three studies specifically examining Xu% were found in the
review of the literature.
Gacono, Meloy and Bridges (2000) have reported Rorschach
findings concerning psychopaths, pedophiles, and persons having
committed sexual homicide. The Xu% means range from .23 to .27,
while the X-% means range from .22 to .26. Substantial X-% values,
especially those of .30 or higher, signify an increase in the probability of
inappropriate behavior.
Rouslin (1997) included Xu% in his Rorshach study of a group of
eating disordered women (n=36) and compared them to nonpsychiatric
and clinical samples. He did not find any statistically significant
differences in their responses as hypothesized. However, Pinto (1999)
found a significantly greater amount of Xu% responses in her study of
50 aggressive and non-aggressive adolescent males. Her subjects, aged
thirteen through fifteen, had been identified for disruptive behavior and
were compared with Exner’s normative, non-clinical sample. Her sample
had an average number of unusual responses of 8.67 and the Exner
sample had a mean of 3.47 unusual responses. Half of all the responses
made by Pinto’s sample were identified as unusual or distinctive. These
results suggest that Xu% is capable of detecting aggressive tendencies
amongst adolescent males.
In an effort to explore common characteristics of seven, self-
described psychic channelers, Dawson (1997) found an Xu% mean
ranging from .26 to .50. The X+% mean for her group ranged from .37
Review of Literature
69
to .67. The results were attributed to her subjects' creativity, abilities to
accomodate to regression, and likelihood to be unconventional.
However, this small, non-clinical sample could also be prone to
diagnostic interpretations indicating antisocial or non-conformist
tendencies, and worse, poor reality testing when considering Xu% and
X+% alone. When Dawson included their Popular mean responses (over
8), the clinical picture revealed that the channelers were able to respond
in a conventional manner.
3. Distorted Form (X-%):
Distorted Form (X-%) represents the percentage of responses that
disregard more appropriate features of blot contours. It tends to
represent a ‘disregard for, or distortion of, reality’ and they occur in
almost every record’. When this percentage is high (.15 to .20) and
associated with a low amount of responses (14 to 16) the administrator
should not be casual about the dysfunction (Exner, 2003).
Distorted Form (X-%) symbolize the percentage of responses that
explain a distortion with reality testing. Weiner (1986) stated people
with numerous minus responses do not accurately perceive the world
as most people do and linked many minus responses to schizophrenia.
Scoring is derived by dividing the sum of Form Quality responses not
commensurate with the blot features (FQx-) by the number of total
responses (Exner, 2003). These are very uncommon responses that
violate the contours of the inkblot. For examiners, these responses can
be very difficult to see, and frequently impossible to locate. Exner’s
Review of Literature
70
nonpatient sample possessed minus responses, but the occurrence was
relatively infrequent in comparison to those of the schizophrenic
reference sample. Exner’s (2003) nonpatients displayed an X-% mean of
.07 and in the schizophrenic sample (Lambda less than 1.0) the X-%
mean was .36. For the schizophrenic sample with a Lambda higher
than .99 the X-% mean was .38, or over five standard deviations higher
than the nonpatient sample. Exner (2003) stated an X-% between .15
and .20 is usually sufficient to indicate problems in cognitive mediation
and the individual is likely to be exhibiting reality testing impairment.
Typically, an average number of minus responses is expected to be from
one to three. Records with three to four minus responses can yield an
X-% mean from .15 to .20. This appears to be consistent with his 1986
study comparing individuals diagnosed with schizophrenia, schizotypal
or borderline personality disorders. The eventual findings for the three
groups suggested that all had some difficulty in their cognitive
mediational functioning. However, the mean X-% for the borderline
sample was .13, .18 for the schizotypal group, and .31 for the
schizophrenic group. The schizophrenic sample averaged more than six
minus answers in their records.
Archer and Gordon (1988) found a similar response style in their
study amongst 134 adolescent inpatients when comparing those who
had been diagnosed with schizophrenia against those who had been
diagnosed as depressed. Both groups displayed X-% means that would
satify Exner’s indicator of poor cognitive mediation, However, the
schizophrenic group displayed a significant mean of .34.
Review of Literature
71
Fong-Hartsfield(2000) included in her Rorshach study sexual
offenders who had been diagnosed with schizophrenia and committed
rape with offenders who had been diagnosed with schizophrenia but
had no sexual component in their crime. The schizophrenic rapists
displayed a statistically significant X-% mean of .29 and the
schizophrenics with nonsexual crimes had a mean of .17. These results
are consistent with her hypothesis that schizophrenic rapists have
poorer reality testing than their low violence counterparts.
Exner, et. al (1975) studied 25 individuals prior to elective
surgery and post-surgery, along with individuals who had experienced
considerable physical problems. They found an elevated X-%, when in
combination with an Anatomy or X-ray response. The results of all of
these studies seem to imply that difficulties in perception and mediation
can be detected by X-% and that stressors such as physical health
problems can contribute to this variable being heightened.
Mediational distortion is reflected by X-% , which represents the
proportion of uncommon responses that disregard the appropriate use
of the blot contours. The objects specified are, difficult to see and, in
many instances, impossible to find. In effect, they are violations of
reality. The X-% also has reasonably substantial long term and short
term retest reliability, yielding coefficients ranging from .80 to .90.
Minus responses are not uncommon, but usually occur in low
frequencies.
Exner(2001) studied that about 86% of 600 nonpatients gave at
least one minus answer, and the mean X-% for the group is
Review of Literature
72
.07(SD=.05). The mean for outpatients is .16(SD=1.0), for inpatient
affective disorder is .19(SD=.10),and for inpatient schizophrenics is
.36(Sd=.13). A substantial proportion of minus answers given by
persons who have significant affective problems include achromatic or
chromatic color or shading determinants (Exner,1993).
Persons with more diffuse cognitive impairment will often give
numerous minus answers that have little homogeneity. Epstein(1998)
has reported that persons with mild or moderate to severe traumatic
brain injury have noticeably low X+%’s and high X-%’s.
Pinto (1999) found that the Rorschach of 50 adolescents with
substantial histories of disruptive behavior contained significantly
higher frequencies of minus and unusual responses.
Bartell and Solanto (1995) have noted higher than expected X-%’s
among children diagnosed with ADHD. Lipgar and Wachler (1991) have
reported that mothers of behaviorally disrupted infants give significantly
fewer from appropriate responses than non patients.
Bannatyne, Gacono and Greene,(1999) studied patterns of
responding on the Rorschach and MMPI-2 for three groups chronic
psychotic forensic patients (undifferentiated schizophrenia,
schizoaffective, and paranoid schizophrenia). They note X+% of less
than .50 in more than half of the records in each group and X-%’s
greater than .20 in two-thirds or more of the protocol in each group.
4. Popular Responses (P):
Rorschach did not mention Popular responses in his monograph,
but did describe them in the posthumously published 1923 paper
Review of Literature
73
(Rorschach & Oberholzer, 1923). He defined them as answers occurring
at least once in every three records, and suggested that they reflect the
ability to perceive and respond to the commonplace features of the
figures. Baughman (1954) found P to be one of the most stable features
in the test, and least subject to undue sets created by examiners. Short-
term reliabilities range from .84 to .88 and over longer intervals from
.79 to .86.
The Comprehensive system includes 13 Popular responses(P),
derived from a North American, English-speaking sample. Non patient
adults and outpatient average nearly seven, with medians of six and
modes of six and four respectively. Schizophrenics average nearly five
with a mode of six and a median of five, while depressive average
slightly more than five, with a mediation of five but with a mode of
eight. A low frequency of Popular responses in the record of an adult,
four or less, reflects either an inability or unwillingness of the person to
deliver that which may be the most obvious possible answer. The
correlation between Popular responses and X+% is negligible, -.02 for
100 nonpatients adults (Exner, Viglione, & Gillespie, 1984). Low P may
signal serious cognitive problems, but it also may reflect a more unique
person who does not violate reality, but instead tends to deal with it in
a common, but not highly conventional manner. More than 90% gave
the Popular response to card VIII, whereas only about 35% gave the
Popular response to card II.
Rapport et al. (1946) recommends the scoring of P for responses
occurring once in every four or five records. Beck et al. (1961) identify P
Review of Literature
74
responses as those occurring at least three times as frequently as the
most common answer to a blot, provided that it is given not less than
once by at least 14% of his adult sample. Piotrowski(1957) included
responses given at least once in every four records. Hertz (1970)defines
any answer that occurs once in six protocols as Popular. Popular (P)
responses are answers from the individual that use the most distinct
and obvious elements of the blots. Rorshach did not mention Popular
responses in his body of work, but he did define them as “Vulgar”
responses that occurred at least once in every three records (Rorshach
& Oberholzer, 1923). “Vulgar” is a translation from his European
linguistic influences of Swiss, Russian, and French and interprets as
banal or lacking creativity. He stated that “Vulgar” responses were a
reflection of the respondent’s ability to identify and react to obvious
shapes of blot contours, or see what others see. Eighty years later, this
same theory of the Popular response is upheld by systematizers today
(Exner, 2003).
Schafer (1954) wrote that interpretative knowledge about a
subject’s adaptation, connection with reality, and defenses could be
learned from their Popular responses. In 1993, Exner conducted a
study with 7500 protocols. This sample consisted of 2,500
nonschizophrenic outpatients, 2500 inpatient nonschizophrenics, and
2,500 nonpatient adults. Response frequencies were then generated
through computer tabulations. Those responses that occurred 2,500
times were designated as popular in the comprehensive system. As a
result, 13 Populars emerged (Exner, 1993). Exner (2003) stated that
Review of Literature
75
most people tend to give six to eight Popular responses. Any more such
responses (10+) can indicate the individual is taking a conventional or
“easy route” in their approach or they are being “obsessive”. Protocols
with four or less Popular responses tended to reflect an “inability or
unwillingness” on the part of the subject to deliver the most obvious
answers. Low responses when taken in consideration of the other
Cognitive Mediation variables and the presenting problem can also
allude to cognitive difficulties, or reveal an exceptional individual who
tends to be unconventional. It appears that stages across the lifespan
can also influence the number of Popular responses. Children at five
years of age tend to provide five responses and this steadily increases.
At ten, they begin to provide an amount equal to nonpatient adults.
However, there have been mixed results in studies examining Popular
responses and senior adults (Reichlin, 1984). It has been assumed that
due to cognitive decline, reluctance to take risks, or unfamiliarity with
test-taking that more aged individual’s Popular response productivity
will decline. But, a study conducted by Geertsma (1962) revealed high
Popular response loadings with normal senior subjects. Additionally,
these results were accomplished with a low number of total responses
(R).
5. Reliability, Validity and Norms in relation to Cognitive
Mediations:
Reliability and Validity of Rorschach Inkblot Test is vary in
different region. Following research finding indicates the valuable
information regarding Reliability and Validity.
Review of Literature
76
5.A. Reliability
Exner (1986) has pointed out that there is not one Rorschach but at
least five because of the five major systems for interpretation. Reliability
and validity studies performed on one system could not be generalized
to another. Despite these difficulties, estimates of reliability can be
obtained by referring to meta-analytic reviews by Parker, Hansen, and
Hunsley (1988). Parker et al. analyzed 39 papers using 530 different
statistical procedures. Parker and colleagues concluded that, overall,
the Rorschach can be expected to have reliabilities in the low to middle
0.80s. While developing the Exner Comprehensive System, Exner
(1993), gave particular attention to interscorer reliability. No scoring
category was included unless it achieved a minimum 0.85 level among
different scorers. Test-retest reliabilities were more variable. Retesting of
25 variables over a one-year interval for a nonpatient group produced
reliabilities ranging from 0.26 to 0.91. Retesting for children did not
come close to the same degree of stability as for 33 adults, although
Exner reported this was to be expected, given that children undergo
considerable developmental changes (Exner & Weiner, 1995). Short-
term retesting over 7-day and 3-week intervals for 9- year-olds indicated
an acceptable level of stability with levels for 25 variables ranging from
0.70 to 0.90.
5.B. Validity:
The primary focus of early validity studies for the Rorschach was
to differentiate empirically among different populations, based on: a)
past observations of a particular group’s responses to the Rorschach, b)
Review of Literature
77
the development of norms based on these observations, and c)
comparisons of an individual’s Rorschach responses with these norms.
For example, depressed individuals tend to have very few human
movement responses (Exner & Weiner, 1995; Miller & Hughes, 1995).
General approaches have resulted in a large number of specific scorings
and interpretations, all of which have had various degrees of validation.
According to Groth-Marnat (1997), establishing validity of the
Rorschach has been complicated by the many scoring categories and
quantitative formulas, each of which has varying levels of validity. Some
interpretations have greater validity than others, even within a specific
category. However, the general consensus among several meta-analytic
reviews was that, when rigorous, high-quality studies were analyzed,
validity ranged from 0.40 to 0.50 (Atkinson, 1986; Parker et al., 1988;
Weiner, 1997), making the Rorschach, overall, achieve reliability and
validity levels comparable to the MMPI and MMPI-2 (Meyer, 1996a;
Meyer, 1999; Stricker & Gold, 1999).
The development of Exner’s Comprehensive System was largely
motivated by the deficiencies (and strengths) inherent in each of the
earlier systems. Recently, as a result of there being a greater proportion
of studies that have used the Exner Comprehensive System some
researchers have been less critical of Rorschach validity (Groth-Marnat,
1997; Meyer, 1996b). However, external validity has continued to cause
critical review of the Rorschach (Groth-Marnat, 1997). Groth-Marnat
(1997) noted that the main focus of Rorschach validity studies has been
directed toward determining its ability to discriminate among different
Review of Literature
78
types of populations; less success has been achieved in making
accurate predictions for areas such as response to therapy, academic
achievement, or spontaneous improvement in a clinical condition.
Under ideal conditions, a test such as the Rorschach should not only
infer characteristics regarding the ways in which individuals organize
their perceptions, but also should convert these inferences into
understanding types of relevant behavior. Multicultural theorists
reported that many measures used to assess students of color did not
capture the relevant facets of behavior or functioning among children of
different cultural groups (Lambert & Rowan, 2003; Haynes et al., 1999).
Lambert and Rowan (2003) stated that measurement developers often
included representative samples of individuals from diverse
background; however, they concluded that the representation was
usually insufficient to explore how these measures functioned within
these different populations.
5.C. Norms:
The Rorschach normative data present two challenges for usefulness.
The first is the size of samples at each age is modest. Only three age
groups (9,11, 16) include more than 130 subjects and two age groups
(5, 6) include less than 100 subjects. Exner, Thomas, and Mason (1985)
also added that the stratification process has potentially created a great
deal of heterogeneity in the data for each group.
Secondly, it has also been noted that all subjects were volunteers,
with parental agreement; in many instances, the actual number of
volunteers with parental agreement was considerably less than the
Review of Literature
79
actual number of potential subjects available in a school or group. This
was especially a problem in recruiting children from urban
communities. Exner and Weiner (1995) concurred that generally fewer
than 20% of the children in a class volunteered for the study, and that
percentage was considerably reduced by the failure of many in
obtaining parental consent. Therefore, Exner and Weiner (1995)
regarded the data as representing children who, for some reason were
interested in taking the test, and whose parents supported this.
Therefore, this may have left out parents who might be “suspicious” of
testing situations (such as urban parents of color), or children not
particularly interested assessment, possibly skewing the data in some
way. Exner and Weiner (1995) stated that the normative data generated
from the protocols of 1580 children between the ages of 5-16 was
stratified for geographic distribution and partially stratified for
socioeconomic level.
As for reported cross-cultural utility of the norms, Exner and
Weiner (1995) stated that in general, “most results are inconsequential”
(p. 48). Males and females do not differ for any location or determinant
scoring, except that females in age groups 12, 13, and 14 gave
significantly more Y (Y is used for responses based on the light-dark
features or shading features of the blot), responses than males for the
same ages. Exner and Weiner (1995) indicated a higher Y is indicative of
“hopelessness.” Differences were more marked when SES was
considered. For ages 5 to 11, the combined group of middlelower and
lowest lower SES gave significantly fewer M responses (M is used for
Review of Literature
80
human movement responses) than children of other SES groups
regardless of race, sex, or geography. A lower level of human movement
might be indicative of depression per Exner (2001), Weiner (1997), and
Frank (1993a). Inexplicably there were a few geographic differences, for
example, southwestern and western children 10 to 14 gave more color
answers (C, CF, and FC combined), than children of the same ages from
other areas of the country. In relation to differences in quantity of
responses, Exner and Weiner (1995) suggested that while the test
process is consistent for perceptual-cognitive operations represented in
the structural data, it might also be markedly influenced by cultural
factors. Despite that conclusion Exner & Weiner (1995) thought
establishing normative data for specific cultures, country, or language
would be “unrealistic” (p. 50).
According to Groth-Marnat (1997), combining the results from a
number of studies, the general consensus among well-designed meta-
analytic reviews was that concurrent validity for the Rorschach ranged
from 0.40 to 0.50. This is nearing validity of the Wechsler Intelligence
Scales for Children- 3rd Edition (WISC-III; Wechsler, 1991), that has
concurrent validity for Full Scale IQ scores ranging from 0.65-0.96 with
a median range of 0.83 (Wechsler, 1991). It is also generally comparable
to the concurrent validity reported in the technical manua1 (1997) for
the Wechsler Adult Intelligence Scale-3rd Edition (WAIS-III; Wechsler,
1997)), that was stated to range from 0.60-0.77.
Exner and Weiner (1995) reported another major factor that
might serve to lower validity to be the meaning associated with response
Review of Literature
81
productivity. Response productivity was found to be closely tied to age,
intellectual level, verbal aptitude, and amount of education (Exner &
Weiner, 1995; Frank, 1994). Norms have been provided for different
ages, but the other three factors (IQ, verbal abilities, and education) can
potentially confound the meanings associated with response
productivity.
Rorschach “scores” are really codes that signify the presence of
certain characteristics within the response (Exner, 1988; Weiner, 1997).
Constellations are of particular importance because they combine
scores into meaningful patterns. Current constellations provide
screening information about schizophrenia, depression, suicide
potential, and interpersonal coping among other manifestations (Exner,
2001). Scores are combined to form seven variable clusters based on
the frequency counts, ratios, percentages, and special indications.
These variable clusters involve personality components or functions
such as affective features, capacity for emotional control and stress
tolerance, ideation, information processing, interpersonal and self-
perceptions, and situationally related stress. In the usual application of
the Rorschach, the most significant emphasis is placed on the final
“global” description of the individual, in which the clinician integrates
the results from different parts of the protocol and takes into account
the interrelations of different scores and reference points.
In the Exner Comprehensive System, the Rorschach is considered
primarily a cognitive perceptual task; responses are believed to measure
the way in which individuals normally react to problem solving
Review of Literature
82
situations. Interpretation of the Rorschach using the Exner
Comprehensive System yields a constellation of elements that forms the
structural aspect of the system. Using Exner’s system, certain elevated
or deficient Determinant or Content variables occurring in an
individual’s Rorschach response set are presumed to reflect
characteristic personality traits, especially those signifying pathological
aberrations (Exner, 2001).
Diagnostic interpretation of the Rorschach is based on normative
data originally derived in large part from adult ‘mainstream’ groups.
More recently, the normative data 38 have been updated to include
some children of color (Anastasi, 1998; Groth-Marnat, 1997; Sattler,
1992). The Exner System presents normative data based on 700
nonpatient adults, and 1390 nonpatient children with separate norms
by age, from 5-16 years, as well as groups of adults with psychiatric
problems, adult inpatients with schizophrenia, adults with depression,
and other adult groups. The Exner System normative data included
statistics for 33 different structural variables. Particular profiles can be
used to screen for various psychological disorders such as suicide
potential or “ego-strength”: determinants can be used to represent a
particular way in which an individual perceives stimuli, reflecting some
aspect of cognitive processing.