Performance of Personality Assessment Inventory and Rorschach Indices of Schizophrenia in a Public Psychiatric Hospital.
ABSTRACT The present study investigated the performance of indices of schizophrenia from the Personality Assessment Inventory (PAI-SCZ; L. C. Morey, 1991) and Rorschach (Rorschach SCZI; J. E. Exner, 1993) in a heterogeneous sample of 24 inpatients at a public psychiatric hospital in the southeastern United States. Results indicated modest agreement between the PAI-SCZ and Rorschach SCZI. More important, the PAI-SCZ but not Rorschach SCZI reliably differentiated inpatients with schizophrenic-spectrum diagnoses from inpatients with other psychiatric diagnoses. In settings in which psychotic disorders falling outside the schizophrenic spectrum are common, the PAI-SCZ may be better suited than the Rorschach SCZI to aid in the differential diagnosis of schizophrenia. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
- SourceAvailable from: jyu.fi
- [Show abstract] [Hide abstract]
ABSTRACT: We systematically evaluated the peer-reviewed Rorschach validity literature for the 65 main variables in the popular Comprehensive System (CS). Across 53 meta-analyses examining variables against externally assessed criteria (e.g., observer ratings, psychiatric diagnosis), the mean validity was r = .27 (k = 770) as compared to r = .08 (k = 386) across 42 meta-analyses examining variables against introspectively assessed criteria (e.g., self-report). Using Hemphill's (2003) data-driven guidelines for interpreting the magnitude of assessment effect sizes with only externally assessed criteria, we found 13 variables had excellent support (r ≥ .33, p < .001; ∴ FSN > 50), 17 had good support (r ≥ .21, p < .05, FSN ≥ 10), 10 had modest support (p < .05 and either r ≥ .21, FSN < 10, or r = .15-.20, FSN ≥ 10), 13 had little (p < .05 and either r = < .15 or FSN < 10) or no support (p > .05), and 12 had no construct-relevant validity studies. The variables with the strongest support were largely those that assess cognitive and perceptual processes (e.g., Perceptual-Thinking Index, Synthesized Response); those with the least support tended to be very rare (e.g., Color Projection) or some of the more recently developed scales (e.g., Egocentricity Index, Isolation Index). Our findings are less positive, more nuanced, and more inclusive than those reported in the CS test manual. We discuss study limitations and the implications for research and clinical practice, including the importance of using different methods in order to improve our understanding of people. (PsycINFO Database Record (c) 2012 APA, all rights reserved).Psychological Bulletin 08/2012; · 15.58 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: This study contributes to the ongoing construct validation of the Personality Assessment Inventory (PAI; Morey, 1991 , 2007 ) by identifying nontest life-event correlates of the PAI full scales and subscales in a sample of psychiatric patients. The life-event data used in this study included education, marital status, and employment, as well as a history of suicide attempts, psychiatric hospitalizations, trauma, medical problems, hallucinations, paranoid ideation, drug abuse, alcohol abuse, and arrest. Correlations were calculated to explore the convergent and discriminant validity of the PAI scales relative to the life-event data. The results showed that the majority of the PAI scales (11 of 13) had meaningful correlations with at least 1 life-event variable. The PAI BOR scale had the greatest number of correlations and was associated with 8 life-event variables. In contrast, the PAI ANX and MAN scales had no correlations above a predetermined threshold (r ≥.21). These findings add to the growing body of empirical correlates of the PAI and generally provide support for the construct validity of the PAI scales.Journal of Personality Assessment 05/2012; 94(6):593-600. · 1.29 Impact Factor
Performance of Personality Assessment Inventory and
Rorschach Indices of Schizophrenia in a Public
E. David Klonsky
University of Virginia
The present study investigated the performance of indices of schizophrenia from the Personality
Assessment Inventory (PAI-SCZ; L. C. Morey, 1991) and Rorschach (Rorschach SCZI; J. E.
Exner, 1993) in a heterogeneous sample of 24 inpatients at a public psychiatric hospital in the
southeastern United States. Results indicated modest agreement between the PAI-SCZ and Ror-
schach SCZI. More important, the PAI-SCZ but not Rorschach SCZI reliably differentiated
inpatients with schizophrenic-spectrum diagnoses from inpatients with other psychiatric diag-
noses. In settings in which psychotic disorders falling outside the schizophrenic spectrum are
common, the PAI-SCZ may be better suited than the Rorschach SCZI to aid in the differential
diagnosis of schizophrenia.
Diagnosing schizophrenia can be challenging.
There is no chemical or anatomical marker by which
a patient can be definitively identified as having
schizophrenia (Gottesman, 1991). Instead, clinicians
make diagnoses according to the presence of a mini-
mum number of symptoms characteristic of schizo-
phrenia (e.g., delusions, hallucinations, disorganized
speech, social/occupational dysfunction) and the ab-
sence of other disorders that could account for such
symptoms (e.g., mood disorders, medical conditions;
American Psychiatric Association, 2001). The ab-
sence of a distinct neuropathology requires that a
multidimensional assessment of behaviors be con-
ducted to ensure accurate diagnoses (Gottesman,
1991). Even when considering multiple sources of
information (e.g., diagnostic interview, behavioral
observations, family history), clinicians frequently
have difficulty arriving at a diagnosis. Psychological
tests may aid in the differential diagnosis between
schizophrenia and other disorders found in psychiat-
Two tests commonly used in psychiatric hospitals
to aid in differential diagnoses are the Personality
Assessment Inventory (PAI; Morey, 1991) and the
Rorschach Comprehensive System (RCS; Exner,
2000). The PAI is a self-report measure of personal-
ity and psychopathology. It includes a schizophrenia
scale (SCZ) designed to measure symptoms associ-
ated with schizophrenic-spectrum disorders. The
PAI-SCZ has repeatedly been demonstrated to cor-
relate with other, well-validated measures of schizo-
phrenic symptomatology (Fantoni-Salvador & Rog-
ers, 1997; Morey, 1999). In addition, the PAI-SCZ
has been shown to differentiate patients with schizo-
phrenia from nonpatient controls (Boyle & Lennon,
1994). However, it is unclear if the PAI-SCZ can
discriminate between diagnostically distinct groups
of psychiatric patients. For example, in a study of
patients with alcoholism and schizophrenia, the two
groups generated similar scores on the PAI-SCZ
(Boyle & Lennon, 1994).
The RCS is a projective measure of personality
and psychopathology. The RCS includes a schizo-
phrenia index (SZCI1; Exner, 1993) designed to mea-
sure the perceptual and cognitive distortions charac-
teristic of patients with schizophrenia and related
psychotic disorders. This index correlates minimally
with self-report measures of schizophrenic symptom-
I thank Dennis Donat, Jeffrey Phillips, and Sharon
Johnson for their assistance in carrying out this study, and
Alexis Black for her comments on earlier versions of this
This research was supported in part by National Institute
of Mental Health (NIMH) Grant MH51187 and by an
NIMH National Research Service Award.
Correspondence concerning this article should be ad-
dressed to E. David Klonsky, Department of Psychology,
University of Virginia, 102 Gilmer Hall, P.O. Box 400400,
Charlottesville, VA 22904-4400. E-mail: klonsky@
1With the most recent update of the RCS (Exner, 2000),
the SCZI has been revised and renamed the Perceptual-
Thinking Index (PTI). A recent study found that the SCZI
and PTI are substantially correlated (r ? .94), indicating
that the two scales essentially measure the same construct
(Smith, Baity, Knowles, & Hilsenroth, 2001).
2004, Vol. 1, No. 2, 107–110
Copyright 2004 by the Educational Publishing Foundation
atology such as the Minnesota Multiphasic Personal-
ity Inventory (MMPI; Hathaway & McKinley, 1940;
Archer & Gordon, 1988; Meyer, Riethmiller, Brooks,
Benoit, & Handler, 2000). Nevertheless, the Ror-
schach SCZI has repeatedly been shown to differen-
tiate patients with schizophrenia and related psy-
chotic disorders from non-psychotic patient groups
(Jorgensen, Andersen, & Dam, 2000). For example,
patients with schizophrenia but not patients with se-
vere depression generated elevated scores on the Ror-
schach SCZI (Ilonen et al., 1999). Likewise, the
SCZI differentiated patients with psychotic disorders
from patients with personality disorders and partici-
pants in a nonclinical control sample (Hilsenroth,
Fowler, & Padawer, 1998).
The studies described above establish the ability of
the PAI and RCS schizophrenia indices to distinguish
between well-defined groups of patients.2The pres-
ent study seeks to extend our knowledge of the PAI-
SCZ and Rorschach SCZI by examining how these
scales operate in a public psychiatric hospital. In this
type of setting, clinicians do not have the benefit of
choosing among a predetermined and limited number
of diagnostic possibilities. Instead, clinicians in pub-
lic hospitals make differential diagnoses among the
diversity of disorders typically encountered in such
settings. The analyses in the current study were con-
ceived with this type of situation in mind. The present
study examines the relationship of the Rorschach and
PAI schizophrenia indices to each other and to Di-
agnostic and Statistical Manual of Mental Disorders
(4th ed.; DSM–IV; American Psychiatric Associa-
tion, 2001) diagnosis in a heterogeneous sample of
inpatients at a public psychiatric hospital.
Participants were 24 inpatients (12 men, 12
women) admitted to a public psychiatric hospital in
the state of Virginia between 1995 and 2000, who
were administered both the PAI and Rorschach. Pa-
tients with valid PAI and RCS protocols were iden-
tified from an archival search of hospital psychologi-
cal evaluation files. Participants had a mean age of 35
(SD ? 12.4). Twenty participants were Caucasian, 3
African American, and 1 Asian.
Diagnoses were made according to DSM–IV crite-
ria by a multidisciplinary treatment team that in-
cluded a board certified psychiatrist, a licensed clini-
cal psychologist, a licensed clinical social worker,
and a registered psychiatric nurse. All team members
participated in a comprehensive intake evaluation of
symptoms and history. In addition, all team members
had the opportunity to interact with the patient as
well as to observe the patient’s behavior on the ward
prior to arriving at a diagnosis. It is important to note
that participants’ diagnoses were determined by the
treatment team before the PAI and Rorschach were
administered. Consequently, PAI and Rorschach data
did not influence the assignment of diagnoses.
Nine participants were assigned DSM–IV diag-
noses of schizophrenic-spectrum disorders (6 schizo-
phrenia, 3 schizoaffective disorder). Diagnoses of the
remaining 15 patients included (the number of par-
ticipants with each diagnosis is indicated in paren-
theses): bipolar disorder with psychotic features (3),
psychotic disorder not otherwise specified (NOS) (2),
substance dependence (2), substance-induced psy-
chotic disorder (2), alcohol dependence (1), major
depression (1), mood disorder NOS (1), mental dis-
order due to a medical condition (1), psychological
factors affecting a medical condition (1), and delu-
sional disorder (1). Two patients had comorbid Axis
II diagnoses of borderline personality disorder.
The PAI is a 344-item self-report instrument that
requires 50–60 min to administer. The PAI-SCZ is a
24-item scale designed to assess symptomatology rel-
evant to the broad spectrum of schizophrenic disor-
ders (Morey, 1999). Participants’ t scores on the PAI-
SCZ were analyzed in the present study.
The Rorschach is a psychological projective test of
personality that requires approximately 96 min to ad-
minister (Ball, Archer, & Imhof, 1994). The RCS
includes the SCZI, designed to aid in the assessment
of schizophrenia and related disorders (Exner, 1993).
The SCZI is based on form quality, cognitive special
scores, and human movement. Scores on the SCZI
can range from 0 to 6. Participants’ scores on the
SCZI were analyzed in the present study.
The PAI-SCZ and Rorschach SCZI were posi-
tively correlated (r ? .42, p ? .04). A two-tailed
t test revealed that PAI-SCZ t scores were substan-
tially higher for the schizophrenic spectrum sample
(M ? 77.4, SD ? 15.5) than for the nonschizo-
2The present literature review of Rorschach validity is by
design limited only to the SCZI. The validity of many RCS
indices remains controversial (e.g., Lilienfeld, Wood, &
Garb, 2000; Viglione, 1999).
phrenic sample (M ? 58.5, SD ? 13.4), t(22) ?
3.16, p ? .005. Rorschach SCZI scores were not
higher for patients with schizophrenic spectrum dis-
orders (M ? 3.11, SD ? 2.15) than for those with
other psychiatric disorders (M ? 2.93, SD ? 1.62),
t(22) ? 0.23, p ? .82.
For both indices, optimal cutoff points for assign-
ing a diagnosis of schizophrenia were determined,
and overall classification accuracy, sensitivity, speci-
ficity, positive predictive power (PPP), and negative
predictive power (NPP) were calculated. For the PAI-
SCZ, an optimal cutoff point of 67 T accurately clas-
sified 79% of patients. Sensitivity was 78%, speci-
ficity 75%, PPP 70%, and NPP 86%. For the
Rorschach SCZI, an optimal cutoff point of 4 accu-
rately classified 54% of participants. Sensitivity was
44%, specificity 60%, PPP 40%, and NPP 64%.
In this study, I examined the relationship of the
PAI-SCZ and Rorschach SCZI to each other and to
DSM–IV diagnosis in a heterogeneous sample of 24
inpatients at a public psychiatric hospital. Results in-
dicate modest but reliable convergence between the
PAI and Rorschach schizophrenia indices. This result
is somewhat surprising because the relationship be-
tween information obtained from Rorschach and self-
report indices is typically small or negligible (Archer
& Krishnamurthy, 1993; Lilienfeld, Wood, & Garb,
2000). In the limited research on the relationship of
the SZCI to self-report measures (i.e., MMPI Scale
8), correlations were slightly above 0 (r ≈ .10; Archer
& Gordon, 1988; Meyer et al., 2000). The conver-
gence observed in the present study is difficult to
interpret, particularly given the absence of other pub-
lished studies on PAI–Rorschach convergence. Fur-
ther study should address the possibility that the PAI-
SCZ and RCS SCZI are exceptions to the
relationships typically observed between self-report
and Rorschach indices.
It was also found that the PAI-SCZ was related to
participants’ DSM–IV diagnoses. Participants with
schizophrenic spectrum disorders had substantially
higher PAI-SCZ scores than participants with other
diagnoses. Just under 80% of patients were diagnosed
accurately by using an optimal cutoff point. These
findings are consistent with previous research sup-
porting the convergent validity of the PAI-SCZ (e.g.,
Fantoni-Salvador & Rogers, 1997; Morey, 1999).
The present study extends this literature, however, by
suggesting that the PAI can aid in the differential
diagnosis of patients with schizophrenia from among
a heterogeneous group of patients with other disor-
ders. This application of the PAI may better approxi-
mate how psychological tests are typically used in
public psychiatric hospitals.
In contrast, the Rorschach SCZI did not reliably
differentiate patients with schizophrenic spectrum di-
agnoses from those with other diagnoses. Although
many studies have shown the ability of the SCZI to
differentiate patients with schizophrenia from non-
psychotic patient or control groups (Hilsenroth et al.,
1998; Ilonen et al., 1999; Jorgensen et al., 2000), the
SCZI is probably less able to distinguish patients
with schizophrenia from patients with psychotic dis-
orders falling outside the schizophrenic spectrum.
The Rorschach SCZI was designed to measure the
perceptual and cognitive distortions characteristic of
psychotic disorders including but not limited to
schizophrenia (Exner, 2000; Hilsenroth et al., 1998;
Jorgensen et al., 2000). Consequently, patients with
psychotic disorders falling outside the schizophrenic
spectrum may generate scores on the SCZI similar to
those of patients with schizophrenic-spectrum disor-
ders. For example, Ilonen et al. (1999) reported that
43% of a sample of manic patients generated elevated
scores on the SCZI. As is typical in public psychiatric
hospitals, most participants in the present study who
did not have schizophrenic spectrum disorders had
disorders associated with psychotic features (e.g.,
psychotic disorder NOS, substance induced psy-
chotic disorder, delusional disorder, bipolar disorder
with psychotic features, severe mood disorders, poly-
substance dependence). That these patients generated
scores on the Rorschach SCZI similar to those of the
patients with schizophrenic-spectrum disorders does
not refute the construct validity of the SCZI. How-
ever, when the differential diagnosis of schizophrenia
from other psychotic disorders is of primary concern,
the PAI-SCZ may be better suited to the task than the
American Psychiatric Association. (2001). Diagnostic and
statistical manual of mental disorders (4th ed. text ed.).
Washington, DC: Author.
Archer, R. P., & Gordon, R. A. (1988). MMPI and Ror-
schach indices of schizophrenic and depressive diagnoses
among adolescent inpatients. Journal of Personality As-
sessment, 52, 276–287.
Archer, R. P., & Krishnamurthy, R. (1993). A review of
MMPI and Rorschach interrelationships in adult samples.
Journal of Personality Assessment, 61, 277–293.
Ball, J. D., Archer, R. P., & Imhof, E. A. (1994). Time
requirements of psychological testing: A survey of
practitioners. Journal of Personality Assessment, 63,
PAI AND RORSCHACH 109
Boyle, G. J., & Lennon, T. J. (1994). Examination of the
reliability and validity of the Personality Assessment In-
ventory. Journal of Psychopathology and Behavioral
Assessment, 16, 173–187.
Exner, J. E., Jr. (1993). The Rorschach: A comprehensive
system: Vol. 1. Basic foundations (3rd ed.). New York:
Exner, J. E., Jr. (2000). A primer for Rorschach interpre-
tation. Asheville, NC: Rorschach Workshops.
Fantoni-Salvador, P., & Rogers, R. (1997). Spanish versions
of the MMPI-2 and PAI: An investigation of concurrent
validity with Hispanic patients. Assessment, 4, 29–39.
Gottesman, I. I. (1991). Schizophrenia genesis. New York:
Hathaway, S. R., & McKinley, J. C. (1940). A multiphasic
personality schedule (Minnesota): I. Construction of the
schedule. Journal of Psychology, 10, 249–254.
Hilsenroth, M. J., Fowler, J. C., & Padawer, J. R. (1998).
The Rorschach schizophrenia index (SCZI): An exami-
nation of reliability, validity, and diagnostic efficiency.
Journal of Personality Assessment, 70, 514–534.
Ilonen, T., Taiminen, T., Karlsson, H., Lauerma, H., Lei-
nonen, K., Walleinius, E., et al. (1999). Diagnostic effi-
ciency of the Rorschach schizophrenia and depression
indices in identifying first-episode schizophrenia and se-
vere depression. Psychiatry Research, 87, 183–192.
Jorgensen, K., Andersen, T. J., & Dam, H. (2000). The
diagnostic efficiency of the Rorschach Depression Index
and the Schizophrenia Index: A review. Assessment, 7,
Lilienfeld, S. O., Wood, J. M., & Garb, H. N. (2000). The
scientific status of projective techniques. Psychological
Science in the Public Interest, 1, 27–66.
Meyer, G. J., Riethmiller, R. J., Brooks, R. D., Benoit,
W. A., & Handler, L. (2000). A replication of Rorschach
and MMPI-2 convergent validity. Journal of Personality
Assessment, 74, 175–215.
Morey, L. C. (1991). Personality Assessment Inventory:
Professional manual. Odessa, FL: Psychological Assess-
Morey, L. C. (1999). Personality Assessment Inventory. In
M. E. Mauish (Ed.), The use of psychological testing for
treatment planning and outcomes assessment (2nd ed.,
pp. 1083–1121). Mahwah, NJ: Erlbaum.
Smith, S. R., Baity, M. R., Knowles, E. S., & Hilsenroth,
M. J. (2001). Assessment of disordered thinking in chil-
dren and adolescents: The Rorschach Perceptual-
Thinking Index. Journal of Personality Assessment, 77,
Viglione, D. J. (1999). A review of recent research ad-
dressing the utility of the Rorschach. Psychological
Assessment, 11, 251–265.