You are on page 1of 19

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/8658395

Methodological Issues in Cross-Cultural and Multicultural Rorschach Research

Article  in  Journal of Personality Assessment · May 2004


DOI: 10.1207/s15327752jpa8202_7 · Source: PubMed

CITATIONS READS
39 763

2 authors, including:

James Allen
University of Minnesota Duluth
91 PUBLICATIONS   2,741 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Intervention Research to Improve Native American Health (IRINAH) View project

Acculturaltion and parenting in the context of trauma and exile: Vietnamese refugees after 25 years in exile View project

All content following this page was uploaded by James Allen on 08 February 2022.

The user has requested enhancement of the downloaded file.


JOURNAL OF PERSONALITY ASSESSMENT, 82(2), 189–206
Copyright © 2004, Lawrence Erlbaum Associates, Inc.

Methodological Issues in Cross-Cultural


and Multicultural Rorschach Research CROSS-CULTURAL
ALLEN
RORSCHACH
AND DANA
METHODOLOGY

James Allen
Department of Psychology
University of Alaska Fairbanks and
Psychosocial Centre for Refugees
University of Oslo

Richard H. Dana
Regional Research Institute for Human Services
Portland State University

Hermann Rorschach researched the utility of his inkblot experiment to understand


psychopathology and cultural differences. Contemporary research with the Rorschach has
evaluated its utility as a test, although it may more properly represent a clinical method with
somewhat different evaluation criteria. Recent controversy regarding the adequacy of the Ror-
schach as a test and the adequacy of its normative data has at times distorted and oversimplified
important methodological issues inherent in the study of cultural difference. Cultural processes
remain a central and inadequately examined variable in Rorschach research; an important
emergent area of inquiry is the Rorschach’s clinical utility as a cross-cultural assessment instru-
ment. We review multicultural and cross-cultural methodological issues intrinsic to contempo-
rary Rorschach research here. Consideration of cultural issues enlarges and enriches the Ror-
schach clinical utility debate and suggests underexplored research strategies that can contribute
to its resolution.

Psychologists in the United States have made extensive use of 412 clinical psychologists reported use of the Rorschach
of the Rorschach in understanding personality and diagnos- in their work. Other surveys have indicated that 80% of
ing psychopathology for more than half a century. The Com- American Psychological Association approved clinical psy-
prehensive System (CS; Exner, 1993) for administration and chology programs provided Rorschach training and 75% of
scoring transformed the Rorschach into an assessment instru- students learned the CS (Hilsenroth & Handler, 1995;
ment for which reliability and validity could be empirically Ritzler & Alter, 1986). The Rorschach was the fourth most
examined using many of the fundamental tenets of classical widely used instrument in a survey of 1,500 clinical psy-
psychometric theory. High-inference psychoanalytic inter- chologists (Camara, Nathan, & Puente, 2000). Interna-
pretation was gradually eschewed in favor of low-inference tionally, the Rorschach enjoys a similar high profile. In a
interpretation. Widespread acceptance of the CS occurred survey of psychologists in Spain, Portugal, and Latin
within a larger professional objective in clinical psychology America (Muniz, Prieto, Almeida, & Bartram, 1999), the
of establishing an empirical basis for practice. The CS em- Rorschach was the third most frequently used assessment
phasizes reliable scoring procedures and development of a instrument, following translated versions of the Wechsler
body of validity data to support interpretive rules that use Adult Intelligence Scale (Wechsler, 1997) and the Minne-
normative data. sota Multiphasic Personality Inventory–2 (MMPI–2;
The CS has been largely responsible for maintaining the Butcher, Dahlstrom, Graham, Tellegen, & Kaemmer,
prominence of the Rorschach in psychological assessment 1989). There is significant international clinical use and re-
practice in the United States. In one recent survey search with the Rorschach (Butcher, Nezami, & Exner,
(Watkins, Campbell, Nieberding, & Hallmark, 1995), 80% 1998); the International Rorschach Society has over 2,400
190 ALLEN AND DANA
members in 23 countries (Weiner, 1999), and over 7,000 tural clinical utility research with the CS.1 Our goal in this ar-
Rorschach research articles have been published (Butcher ticle is to encourage the development of an empirical basis
& Rouse, 1996). for cross-cultural and multicultural practice with the Ror-
schach through development of a methodologically sound
body of research testing its cross-cultural and multicultural
CS CONTROVERSY clinical utility. We begin our discussion with a review of the
current controversy regarding the cross-cultural adequacy of
Despite continued widespread acceptance of the Rorschach CS norms, which has important relevance to Rorschach re-
by professional psychologists, as evidenced by this survey search and practice.
data, controversy concerning its clinical use has existed for
over 50 years. The earliest Rorschach critiques have exam-
ined shortcomings in its psychometric foundations and valid- VALIDITY CONSIDERATIONS RELEVANT
ity research (e.g., Cronbach, 1949; Eysenck, 1959; Jensen, TO CROSS-CULTURAL AND MULTICULTURAL
1965; McFarlane, 1942; Mensh, 1950; Munroe, 1945). In his RORSCHACH RESEARCH AND PRACTICE
seminal critique, Cronbach described inadequacies of the
Rorschach as a psychometric instrument. These inadequa- A key consideration is typically glossed over by both sides of
cies, along with statistical errors of omission and commis- the CS debate. Confusion is created when authors question
sion, drew into question the conclusions in over 90% of the the validity of the Rorschach or the CS, a point now acknowl-
Rorschach studies Cronbach reviewed. Contemporary criti- edged by one group of Rorschach antagonists (Garb, Wood,
cism reiterates many of these same still unresolved issues Lilienfeld, & Nezworski, 2002). Underlying this question are
(e.g., Dawes, 1999; Garb, Wood, Lilienfeld, & Nezworski, two fundamental misunderstandings. First, on multiple in-
2002; Hunsley & Bailey, 1999; Lilienfeld, Wood, & Garb, dexes instruments such as the Rorschach, the validity status
2000; Wood & Lilienfeld, 1999; Wood, Lilienfeld, Garb, & associated with individual indexes and variables differs. Dis-
Nezworski, 2000a, 2000b; Wood, Lilienfeld, Nezworski, & cussion of validity requires reference to the interpretations
Garb, 2001; Wood, Nezworski, Garb, & Lilienfeld, 2001a; associated with data from a specific CS variable or index.
Wood, Nezworski, & Stejskal, 1996; Wood, Nezworski, Even more important from our perspective, recent contro-
Stejskal, & Garven, 2001), and numerous responses have ap- versy has generally ignored a fundamental tenet of validity
peared in the literature to address these critiques (Bornstein, theory. Empirical research does not validate a test but instead
2001; Gacono, Loving, & Bodholdt, 2001; Ganellen, 2001; validates the interpretations derived from test data
Hibbard, 2003; Meyer, 2001; Miller, Rosenthal, Bornstein, (Cronbach, 1971); validity is not a property of a test or index
Berry, & Brunell-Neuleib, 1999; Stricker & Gold, 1999; of a test.
Viglione, 1999; Weiner, Spielberger, & Abeles, 2002). The A point of departure for this article, which concerns itself
controversy diminishes not only Rorschach credibility; it with study of the cross-cultural and multicultural clinical
threatens acceptance of all standard assessment instruments utility of the CS, is a corollary of this basic tenet of validity
among third-party reimbursers and service providers. theory; validity is not a general characteristic of a test but in-
Piotrowski (1999) and Stout (1997) both have noted a trend stead specific to a particular test purpose (Dana, 1962;
of diminished use of standard assessment instruments, in par- Wainer & Braun, 1988). Thus, interpretations from a specific
ticular in managed care practice settings, despite a recent test can be highly valid for one purpose but not for other pur-
meta-analysis using 800 studies that reported validity coeffi- poses; interpretations may be highly valid for one ethnic or
cients for the Rorschach and other standard psychological cultural population and not valid for another cultural setting.
tests that matched or exceeded those of standard medical Validation of interpretations from CS data in different cul-
tests (Meyer et al., 2001). tural settings is the core concern of research that seeks to de-
In this article, we seek to expand the scope of response to velop an empirical basis for its cross-cultural and
CS antagonists, along with the scope of its clinical applica- multicultural clinical use.
tion, by (a) reviewing findings from recent multicultural re-
search with the CS, (b) discussing methodological issues in Cross-Cultural and Multicultural Rorschach
this work, and (c) describing a research agenda for future Clinical Application
work. We seek to foster Rorschach credibility by enhancing
the methodological quality of cross-cultural and multicul- A recently emerging issue in the Rorschach controversy is al-
leged “shortcomings of the Rorschach for the assessment of
1
American minority groups and for non-Americans” (Garb,
We refer here to application of the Rorschach across cultural Wood, Nezworski, Grove, & Stejskal, 2001, p. 436; see also
groups, often internationally, as cross-cultural application. We refer
to application with ethnic minorities within a specific nation as Lilienfeld et al., 2000). In support of their assertion, Garb et
multicultural application, recognizing the frequent overlap inherent al. cited findings of 18 authors. In the process, Garb et al.
in cultural and minority group status. misconstrued these authors’ intentions and the meaning of
CROSS-CULTURAL RORSCHACH METHODOLOGY 191
their findings through oversimplification, omission of rele- specific procedures. This, in part, has led to allegations of
vant content, and an apparent failure to understand the meth- “cultural negligence” on the part of mainstream clinical psy-
odologies, intent, and findings of the cross-cultural and chology (Kazarian & Evans, 1998).
multicultural assessment literature. This provides little in the Although we believe it is appropriate to criticize the cur-
way of resolution of the Rorschach controversy as alluded to rent research base of the Rorschach for cross-cultural prac-
in the title of Garb et al.’s article. Of greater concern, the arti- tice as deficient, we also believe selective targeting of the
cle obfuscates the intentions and accomplishments of current Rorschach (e.g., Wood, Garb, Lilienfeld, & Nezworski,
research efforts to develop an empirical basis for 2002) is unwarranted and ignores the existence of similar de-
cross-cultural and multicultural Rorschach practice (e.g., ficiencies in the research basis for cross-cultural practice of
Dana, 1993, 2000b). most other standard personality assessment instruments.
In a specific example, Garb et al. (2001) cited a quote from This state of affairs can only occur within a professional con-
Dana (1993, p. 160) indicating that “the Rorschach and the text that possesses limited cross-cultural and multicultural
Exner Comprehensive versions are not recommended for rou- research to guide practice, limited understanding of the
tine cross-cultural application” (p. 436; italics added). The methodological issues inherent in this type of cross-cultural
complete quotation, however, refers to a preceding table of research, and limited awareness of the relatively new consen-
comparisons across major projective methods indicating the sus on multicultural research standards (e.g., Council of Na-
status of culture-relevant desiderata including stimuli, scores, tional Psychological Associations for the Advancement of
norms, context, and theory for several projective tests/meth- Ethnic Minority Interests [CNPAAEMI], 2000).
ods (see Dana, 1993, p. 148, Table 8–1). The complete quota- As Widiger (2001) suggested, “the dispute over the Ror-
tion also indicates that there were “instances with acculturated schach may be a flashpoint of a wider and more fundamental
assesses and for particular assessment problems where either dispute within the profession” (p. 374), described further in
the original Rorschach or the Exner Comprehensive Ror- Howard et al. (1987). We also view this dispute as a flashpoint
schach should be useful” (Dana, 1993, pp. 160–161). Each of but in a way additional to that envisioned by Howard et al.; it
the desiderata in the Dana (1993) Table 8–1 indicated an area incorporates the debate over whether culture is properly con-
in which additional research is needed to document CS valid- sidered as a proximal or distal variable in research and prac-
ity. tice. In their review of ethnic minority assessment research,
The new American Psychological Association (2002) Eth- Okazaki and Sue (1995) organized approaches by the proxim-
ics Code includes a number of newly introduced subsections ity of cultural variables in study design and interpretation to
under the area of assessment that are particularly relevant to central validity considerations;2 as an example of this distinc-
multicultural practice. It calls for use of instruments with es- tion, they noted that ethnicity operationalized as a demo-
tablished reliability and validity among members of the popu- graphic variable places culture more distally in that ethnicity
lation tested or discussion of limitations in the case where they as a demographic variable often functions as proxy for the ac-
do not exist, and it mandates test interpretation to consider lin- tual culturally based personality processes of interest that un-
guistic and cultural differences. It also asserts assessors may
adapt instruments “in light of the research on or evidence of”
2
usefulness (American Psychological Association, 2002, p. The proximal/distal is a theme that appears throughout the multi-
1070). However, routine cross-cultural application of any in- cultural research and practice assessment literature. In their assess-
ment practice recommendations, Sue and Zane (1987) stressed the
strument, including the Rorschach, follows the establishment importance of proximal linkages that transform cultural knowledge
of standards for multicultural and cross-cultural use of the test into concrete operations in practice. Examples of proximal linkages
in assessment practice. Thus, clinical competency in include use of culture-specific assessment service delivery with re-
cross-cultural application is predicated on adherence to proce- spect to cultural norms of helping behavior and social etiquette and
dures relative to standards (Beutler, Crago, & Arizmendi, the use of information on acculturation status as moderator variables
in assessment interpretation. Similarly, the Multicultural Assess-
1986). The new American Psychological Association Ethics ment-Intervention Process model (Dana, 2000b) provides a number
Code now appears to recognize cultural adaptations derived of specified opportunities to employ cultural knowledge within the
from cultural knowledge and clinical judgment in the adminis- assessment process per se or within a community mental health cen-
tration or interpretation of standard tests. However, in the ab- ter service delivery system (Dana, Aragon, & Kramer, 2002). Re-
sence of a consensually accepted research basis for such search-relevant examples following Sue and Zane’s recommenda-
tions include Malgady’s (1996) suggestions for reversal of the null
adaptation, standards for culturally competent assessment hypothesis of no cultural bias to that of culture bias, thereby revers-
practice cannot exist. The absence of a body of acceptable ing Type I and II error with regard to cultural difference based on the
cross-cultural research for major assessment instruments, in- consequences of potential disservice to ethnic minority clients;
cluding but not limited to the Rorschach, on the previously Okazaki and Sue’s (1995) assessment methodology considerations
mentioned desiderata point to areas of need for additional re- regarding the specificity of the ethnicity variable through adequate
description of the sample, or preferably through direct measure of
search as a basis for practice. In this way, a paucity of equiva- personality processes thought to affect test performance for which
lence research has been responsible for the absence of research ethnicity functions as proxy variable, is yet another example of how
based, cross-cultural/multicultural assessment standards and cultural variables become proximal to interpretation of the data.
192 ALLEN AND DANA
derlie ethnicity, which in turn leads to a lack of specificity in underlie the ethnic group status of a number of the
the independent variable. Proximal loci and specificity of the participants. These considerations emphasize an unintended
cultural variable are central methodological considerations in consequence of Wood, Nezworski, Garb, et al.’s cultural di-
cross-cultural and multicultural research with the Rorschach versity critique of the Rorschach. Given the methodological
as well as all other assessment instruments. This lack of speci- shortcomings of Wood, Nezworski, Garb, et al.’s study, an
ficity in the research has resulted in unsystematic introduction alternative interpretation of their findings highlights unique
of numerous variables associated with culture, along with “all positive characteristics of the Rorschach as a cross-cultural
manner of demographic variables such as age, sex, regional lo- instrument, tapping potentially important cultural differ-
cation, interest, acculturation, [and] socioeconomic status” ences in central personality variables among ethnically di-
(Hibbard, 2003, p. 262) that can potentially mediate observed verse samples. We argue in the following that the
relationships and differences. This methodological flaw is Rorschach’s unique potential as a cross-cultural assessment
common in much of the current multicultural assessment re- instrument offers a powerful rationale for its continued use,
search, rendering clear conclusions problematic; it has partic- particularly internationally, as well as an avenue of construc-
ular relevance to cross-cultural comparisons of CS scores to tive response to CS criticisms. This has important implica-
the CS norms. tions for the study of the clinical utility of the Rorschach and
other assessment instruments.
Adequacy of CS Norms
CROSS-CULTURAL/MULTICULTURAL
Wood, Nezworski, Garb, et al. (2001a) assembled a large sam- CLINICAL UTILITY OF THE RORSCHACH
ple combining data from 32 Rorschach studies of nonclinical
populations. Wood, Nezworski, Garb, et al.’s (2001a) compar- An important issue in the CS clinical utility debate generally
isons of the composite means of 14 CS variables from these omitted in the literature is consideration of its cross-cultural
combined samples to the current CS norms yielded statisti- and multicultural applicability. This issue is now frequently
cally significant differences. Based on this data, Wood, addressed with many other widely used psychological tests
Nezworski, Garb, et al. (2001b) asserted the current CS norms (e.g., Dana, 2000b; Lonner, 1996; Roysircar-Sodowsky &
are flawed, despite later observing, “the numbers in our article Kuo, 2001; Sue, Kurasaki, & Srinivaran, 1999). Given cur-
are unsuitable for norms” (p. 398). Meyer (2001) responded by rent demographic transitions in the United States and glob-
citing numerous sampling considerations and analytic errors ally, consideration of cross-cultural utility in the evaluation
that limit the interpretability of the analysis of data from of clinical utility is inevitable (Dana, 2000a; Sanz, Ritzler,
Wood, Nezworski, Garb, et al.’s (2001a) composite sample. Ivanouw, & Dana, 2001).
Additional considerations associated with multicultural Hermann Rorschach’s original 1921 monograph is note-
sampling concerns inherent in the Wood, Nezworski, Garb, worthy in its description of the inkblot procedure as a useful
et al. (2001a) data expand Meyer’s (2001) critique. Contra- technique for investigation of psychopathology as well as cul-
vening basic methodological guidelines for ethnic minority tural variation in personality (Rorschach, 1921/1942). Ror-
research (Okazaki & Sue, 1995; CNPAAEMI, 2000), the schach contrasted cultural differences in experience type
Wood, Nezworski, Garb, et al. composite data set was cre- between Bernese and Appenzeller canton residents and find-
ated through nonsystematic inclusion of African American ing it interacted with differences in the kinds and frequencies
and Hispanic samples without description of their cultural of psychopathological disorders and specific
and ethnic characteristics. Although Meyer’s (2001) inter- symptomatologies. Rorschachs’s (1921/1942) statement that
pretation of differences as representative of normative shifts “the experience type should be very different in various people
in the contemporary samples are reasonable, consideration of and races” (p. 96) represents the original impetus for studies of
the ethnic composition of the Wood, Nezworski, Garb, et al. Rorschach cross-cultural/multicultural utility and suggested
samples expands the parameters of the debate regarding the hypotheses for factor analytic investigation of the experience
CS norms to cultural factors. The Wood, Nezworski, Garb, et balance (eb) as a bona fide etic or cross-culturally universal
al. findings could be due, at least in part, to the effects of their psychological construct (Dana, Hinman, & Bolton, 1977).
nonsystematic inclusion of ethnic minority participants. Lindzey (1954) noted that Hermann Rorschach provided
More broadly, any potential CS normative shifts could be this preliminary data on group cultural differences because
grounded, in part, in recent cultural and demographic transi- he recognized his inkblot test’s unique potential for
tions in the U.S. population. Because of their neglect of im- cross-cultural transportability given its composition from
portant multicultural methodological considerations relevant task stimuli of neutral, nonpictorial, visual forms requiring
to their inclusion of ethnic minority data, the design of the no item translation. This potential remains underutilized to
Wood, Nezworski, Garb, et al. study does not allow the re- this day; the CS provides brief standard instructions, reliable
searchers to conclude that any observed differences in CS coding, norms, and validity data, but the generalizability of
scores, when compared to the existing CS norms, are not in- these CS elements beyond Euro-Americans living in the
stead due to cultural differences tapped by CS variables that United States remains an important empirical question.
CROSS-CULTURAL RORSCHACH METHODOLOGY 193
One frequently voiced concern involves a critique of the ing a time when clinical utility will be increasingly under-
projective hypothesis as culture specific (cf. Draguns, 1990). stood in terms of an instrument’s cross-cultural and multicul-
This critique appears to confuse three distinct issues. First, tural applicability. Important methodological issues arise in
the CS conceptualizes Rorschach scoring as primarily as- research that explores the cross-cultural and multicultural
sessing a perceptual and associational (or informa- clinical utility of the Rorschach.
tion-processing) task (Exner, 1993); projection may or may
not be part of many Rorschach responses, and typical proto- METHODOLOGICAL ISSUES
cols do not necessarily include significant projective materi- IN CROSS-CULTURAL AND MULTICULTURAL
als. Substantive CS interpretative data describing the person RORSCHACH RESEARCH
can occur without consideration of projection. Second, these
critiques generally examine inappropriately applied cul- Core methodological cross-cultural/multicultural issues in
ture-specific interpretations of projection by the examiner Rorschach research include (a) linguistic equivalence, (b)
(Draguns, 1990; Okazaki & Sue, 1995). These critiques re- setting and instructional set equivalence, (c) interrater coding
duce to legitimate critiques of theoretical and interpretative reliability, (d) cross-cultural/multicultural norms, (e) accul-
formulations regarding projective material that are embed- turation status norms, (f) predictor bias, (g)
ded in worldviews incongruent with the cultural context of cross-cultural/multicultural construct equivalence, and (h)
the examinee rather than direct critiques of the projective hy- construct validation research. Descriptions and recommen-
pothesis itself. Third, ethnic match may influence projection; dations for each of these methodological issues are summa-
indeed, issues and implications associated with exam- rized in Table 1. Some of these issues are common to cultural
iner–examinee ethnic match and cultural competence of the research with other assessment instruments; in such cases,
examiner constitute important ramifications of the projective we provide a brief comparative overview of recent findings,
hypothesis unexamined in the assessment literature. most notably with the MMPI–2. Other multicultural method-
In summary, attributes of the Rorschach afford it a unique ological issues are more specific to the Rorschach and not of
potential as a truly international assessment instrument dur- direct relevance to self-report instruments.

TABLE 1
Core Cross-Cultural Methodological Issues

Issue Description Recommendation

Linguistic equivalence Address language translation and Standard procedures to translate instructions including
ethnic/regional differences in English prompts
usage
Setting/instructional set Cross-cultural appropriateness of Empirical study of administration, interactional style, and
equivalence standard administration meaning in terms of cultural sociolinguistic conventions
Empirical study of examiner–examinee ethnic match
Interrater coding reliability Comparability of cross-ethnic and Implement existing recommendations for research studies
cross-national samples along with careful specification of any differences in
cross-national coding procedures
Cross-cultural/multicultural Correct use of cross-national/cross-ethnic Use limited to screening for group differences in CS variable
norms comparisons of norms scores; difference may demonstrate either (a)
nonequivalence of the measure or (b) valid measurement of
cultural difference on underlying construct; only
differential validity studies can address meaning of group
difference scores
Acculturation status norms Establishment of norms for identification Minimize pathologization of individuals with traditional
of cultural orientation of assessee cultural orientations by altering clinician through use of
potentially inappropriate normative data and nonequivalent
constructs
Predictor bias Identification of potential bias in Detection of slope and intercept bias in Rorschach-based
Rorschach prediction predictions of psychopathology and personality (note
implicit assumption of construct equivalence)
Screen for cross-cultural Hypothesis testing research to disconfirm Regression of cultural identity measures on CS variables
construct equivalence null hypothesis of cross-cultural Item Response Theory analysis
nonequivalence of Rorschach variables Confirmatory Factor Analysis
(e.g., Malgady, 1996)
Cross-cultural construct Cross-cultural construct validity research Construct validation studies for CS variables among cultural
validation research groups of interest, including mixed quantitative/qualitative
ethnographic designs
194 ALLEN AND DANA
Linguistic Equivalence tions. Difficulties with self-report instruments and their asso-
ciated instructional sets have been described in use with
Understanding of the current CS instructions as intended by groups ranging from unacculturated non-Western individu-
the CS developers, with their combination of brief initial in- als to counselors completing multicultural competence in-
structions and often complex prompts, may require a particu- ventories (Pope-Davis, Liu, Toporek, & Brittan-Powell,
lar style of abstraction that is itself culture specific. However, 2001). Despite this, instructional set equivalence has not yet
the linguistic equivalence of these instructions has never been systematically studied with standard self-report assess-
been formally studied either for translations (Geisinger, ment instruments such as the MMPI–2.
1994; Van de Vijver & Hambleton, 1996), or equally impor-
tant, for ethnic/regional differences in English usage within Interrater Coding Reliability
and outside of the United States (for an MMPI–2 Singapore
example, see Wuts, 2002). Standard systematic translation Interrater reliability does not arise as a concern for self-report
procedures such as back translation or committee procedures instruments such as the MMPI–2; however, it comprises a
can be readily applied to the entire CS administration process crucial concern in research and clinical application of com-
including the opening instructions and prompts. Following plex scoring systems for verbal responses such as those pro-
this, evaluation through field testing is necessary to examine duced by the Rorschach. Introduction of the CS provided ma-
several important issues, most prominent the applicability of jor improvements in interrater coding reliability among 95%
translation/adaptation by bilinguals for use with monolingual of scoring variables (see review in Weiner et al., 2002). Con-
populations that may not share the same linguistic habits and temporary Rorschach studies using the CS employ clearer
implicit understandings (Nichols, Padilla, & Gomez- coding guidelines and use more sophisticated measures of
Maqueo, 2000). The issue of linguistic equivalence is rele- agreement (Meyer et al., 2002; Viglione & Hilsenroth, 2001)
vant to all standard personality instruments; the MMPI–2 has than pre-CS studies (e.g., Dana, 1955; Voigt & Dana, 1964).
been translated into more than 150 languages using Cross-cultural and especially international research requires
back-translation procedures (Butcher, 1996). However, field careful specification of CS coding procedures, examiner
testing of these MMPI–2 translations has received limited at- training in consensual practice, use of training materials with
tention in the literature. high and low base rate coding scoring, and ongoing
evaluative feedback to examiners and scorers.
Setting/Instructional Set Equivalence Informal reports of international variations in inquiry and
coding suggest that local researchers must clearly specify
As is the case with linguistic equivalence, despite the their procedures. Cultural adaptations may dictate departures
lengthy, intensive one-on-one examiner–assessee interaction from CS procedures in research studies. However, reports
required in CS administration, there is also little research on must (a) identify all such departures from CS procedures, (b)
the cultural appropriateness of the standard CS administra- provide the rationale for departures, (c) discuss potential im-
tion procedure among ethnic minority groups in the United plications, and (d) rename adapted and nonequivalent scor-
States and internationally. Culturally appropriate ing variables. Without these cautions, cross-cultural
interactional styles are directed by sociolinguistic conven- comparisons become hazardous; nonequivalent and contra-
tions. These conventions include such features as appropriate dictory cross-cultural findings can result from the presence
pacing of communication and modes of back channeling. of undocumented CS alterations that distort scoring vari-
Adherence to these conventions is critical in the establish- ables. This distortion can lead to artifactual findings of group
ment of rapport and can affect task orientation, defensive- norm differences, factorial variance, or differential validity
ness, response frequency, and a myriad of associated CS vari- and erroneously suggest the existing nomological net vali-
ables. This is because the sociolinguistic conventions of each dating an underlying construct is not transferable across cul-
culture convey meaning and in addition convey a comfort- tural data sets when instead, changes in scoring procedures
able and credible interactional style for the assessee. Finally, have resulted in variables of interest that are different across
although discussion of its potential importance appears in re- data sets.
cent Rorschach literature (e.g., Presley, Smith, Hilsenroth, &
Exner, 2001), the impact of ethnic match between examiner Cross-Cultural/Multicultural Norms
and examinee during Rorschach administration has not been
explored. Ethnic matching is empirically related to public There are no generally accepted multicultural norms for the
sector managed care treatment outcomes in the United States Rorschach or other standard assessment instruments in the
(Gamst, Dana, Der-Karabetian, & Kramer, 2000, 2001) as United States. Instead, in the case of the MMPI–2, ethnicity
well as to treatment factors and outcome in the psychother- is typically considered one of several demographic variables
apy (Atkinson, 1983; Maki, 1999). stratified through sampling procedures; the resulting com-
Although self-report instruments typically require limited posite normative data is used in interpretation (Butcher et al.,
interaction with the examiner, they involve written instruc- 1989). In the case of the Rorschach, national CS norms for
CROSS-CULTURAL RORSCHACH METHODOLOGY 195
countries outside the United States are not available with the cans score higher on MMPI and MMPI–2 clinical scales
exception of Portugal (Pires, 2000). However, there are a (Frueh, Smith, & Libet, 1996; Graham, 1990; Greene,
number of normative adult samples for the CS from Argen- 1987), leading to continued calls for separate ethnic group
tina, Belgium, Denmark, Finland, Israel, Italy, Japan, Peru, norms (Lindsey, 1998). These findings of differences on
and Spain (Meyer, 2001; T. Schaeffer, personal communica- the MMPI and MMPI–2 scales have often been small, in-
tion, November 3, 2002), and comparable samples from Ven- consistent, and difficult to replicate (e.g., Frueh, Gold, de
ezuela and Chile are cited in Vinet (2000). The availability of Arellano, & Brady, 1997). Hall et al. (1999) found only
this cross-cultural data allows researchers to screen for dif- small effect sizes for the demographic variable of race/eth-
ferences between cultural groups on CS variables similar to nicity across the 37 studies used in their meta-analysis. In
the 25 comparative MMPI and MMPI–2 studies in the United summary, the existence and meaning of ethnic group norm
States that permitted a recent meta-analytic investigation of differences on the MMPI/MMPI–2 remains controversial
ethnicity and the MMPI (Hall, Bansal, & Lopez, 1999). The among MMPI–2 proponents and antagonists.
CS comparative research to date has produced mixed results Resolution to this controversy is found through noting that
regarding cross-cultural differences. problems with these types of between-group comparisons
Wood and Lilienfeld (1999) cited 11 studies in which they have been described for years (e.g., Azibo, 1988; Greene,
asserted U.S. ethnic minorities and non-Americans score dif- 1987; Okazaki & Sue, 1995; Pritchard & Rosenblatt, 1980;
ferently than U.S. Euro-Americans on key Rorschach vari- Roysircar-Sodowsky & Kuo, 2001). The finding of a group
ables. However, only 6 of these studies used the CS, and of difference on a Rorschach, MMPI, or other personality as-
these, 1 study contrasted results qualitatively on only three sessment instrument variable does not permit conclusions re-
variables with the CS norms (Sangro, 1997), whereas the re- garding its meaning. It is unclear whether a between-group
maining 5 studies compared targeted minority convenience difference is due to (a) a real difference in personality or
samples to the Exner normative data. Elsewhere, Wood, psychopathology attributes, (b) nonequivalence of the mea-
Nezworski, Stejskal, Garven, and West (1999) argued such sure or its metric properties, or (c) a shift that has occurred
targeted sample comparisons with the Exner norms are meth- across cultures in the underlying construct tapped by the
odologically flawed. measure and its associated nomological network (Allen &
Meyer (2001) tested for differences among 69 CS variables Walsh, 2000). “Accordingly, such obtained differences have
in samples from Argentina, Belgium, Denmark, Finland, Ja- no necessary implications for the validity of using a test with
pan, Peru, Portugal, and Spain when compared to the CS diverse cultures” (Weiner et al., 2002, p. 8). Research ques-
norms. Meyer found differences cross-culturally; however, tions regarding the cross-cultural appropriateness of the Ror-
these differences averaged only .4 SDs, which he interpreted schach or any psychological test are better addressed through
as statistically but not clinically significant. Wood, methodologies, described following, that explore the relation
Nezworski, Garb, et al. (2001b) reanalyzed this same data set, of test variables among a cultural or ethnic group to (a) a
restricting their analyses to only 14 of the 69 variables they as- cross-culturally equivalent criterion variable or (b) the un-
serted were selected because they were central to CS interpre- derlying nomological net of the variable through tests of con-
tation. Wood, Nezworski, Garb, et al. reported differences on vergent and divergent validity.3
these selected variables averaging .8 SDs, which they inter-
preted as clinically meaningful differences. Acculturation Status Norms
Finally, Ritsher, Slivko-Kolchik, and Oleichik (2001)
found similar means in 104 of 107 Rorschach variables when Investigation of potential multicultural group differences
comparing a Russian inpatient sample with depression to raises accompanying questions regarding ethnic group
Exner’s (1993) U.S. sample of 315 with depression; how- norms and acculturation status norms. The conventional as-
ever, Scale 3 (Depression) from the Russian version of the sessment view militates against establishment of norms for
MMPI but not the CS DEPI scale functioned as a valid indi- all ethnic subgroups because of the within-group heterogene-
cator of depression in this sample. In summary, Rorschach ity of scores. Because of this, the development of local norms
proponents and antagonists disagree on the magnitude, in the United States has been an infrequent recourse limited
meaning, and importance of findings of group difference on primarily to geographically isolated and unacculturated in-
CS variables cross-culturally. digenous groups (Dana, 1993). Neither the CS nor MMPI–2
In the United States, initial between-group comparative provides ethnic group norms.
studies of CS norms have been done with African Ameri-
cans. Presley et al. (2001) found only one clinically signifi-
3
cant difference among 23 core Rorschach variables in a Moreover, despite Malgady’s (1996) cogent critique advocating
comparison between 44 African Americans matched on key the directionality of the null hypothesis in ethnic minority assess-
ment research be set to assumption of cultural difference, given the
demographic variables with 44 Euro-Americans, all se- potential deleterious impact on ethnic minority clients of improper
lected from the CS nonpatient norm group. Several studies test use, most research on group differences tests the hypothesis of
with the MMPI and MMPI–2 have found African Ameri- no group difference as the null.
196 ALLEN AND DANA
The development of acculturation norms is now important Predictor Bias
to consider because of the documented importance of accul-
turation status across ethnic minority groups and the con- Although often described as a technique to identify test bias,
founding of traditional acculturation status with the concept of predictor bias can be understood more broadly
psychopathology (Dana, 2000a).4 Acculturation status repre- from the perspective of basic validity theory, here referring to
sents both an important source of heterogeneity within ethnic the case in which there exists differential validity for a test
groups and a specific variable of interest underlying ethnic score in the prediction of a criterion variable for individuals
group membership. The rationale for acculturation status who are members of a particular ethnic group. In the most
norms is supported by empirical research with the comprehensive cross-ethnic study of predictor bias5 in a per-
MMPI/MMPI–2 among the four major ethnic/racial minority sonality instrument to date, Meyer (2002) explored the rela-
groups in the United States (Dana, 2000a; Whatley, Allen, & tionship of ethnicity to Rorschach variables in a sample of
Dana, 2003). In this research, relationships between accul- 432 Euro-American, African American, Hispanic American,
turation/cultural identity status and MMPI/MMPI–2 clinical Asian American, and American Indian psychiatric patients
scale scores have been reported in studies with Asian Ameri- matched on key demographic variables. Using the approach
cans (Sue, Keefe, Enomoto, Durvasula, & Chao, 1996), His- of Nunnally and Bernstein (1994), Meyer (2002) found no
panic Americans (Montgomery & Orozco, 1985; Whitworth evidence of slope bias on 17 key CS variables in comparisons
& McBlaine, 1993), American Indians (Hoffmann, Dana, & of Euro-Americans to African Americans and of
Bolton, 1985; Pollack & Shore, 1980), and African Ameri- Euro-Americans to a combined group that included all four
cans (Whatley et al., in press). Additionally, acculturation ethnic minority groups. However, 4 of these CS variables, all
status emerges as a performance correlate for Hispanics on associated with psychotic disorder, displayed moderate inter-
standard neuropsychological tests (Arnold, Montgomery, cept bias in the direction of underprediction of
Castanada, & Longoria, 1994). psychopathology among African Americans, and 3 of these 4
The availability of acculturation status norms for the four variables displayed intercept bias, again in the direction of
major ethnic groups in the United States could minimize underprediction of psychopathology, among the composite
pathologization of traditional individuals. By allowing for ethnic minority group.
assessment of acculturative status, clinicians can then iden- Similar research with the MMPI–2 has been limited to Af-
tify traditional individuals for whom the existing normative rican American samples compared to Euro-Americans.
data for the CS and other psychological assessment instru- McNulty, Graham, Ben-Porath, and Stein (1997) and
ments such as the MMPI–2 may be inappropriate as well as Timbrook and Graham (1994) both failed to find evidence of
other ethnic minority individuals for whom standard inter- slope bias in MMPI–2 based predictions of
pretations may be relevant. As a global society with increas- psychopathology. Arbisi, Ben-Porath, and McNulty (2002)
ingly permeable borders, it is now commonplace to find found mean differences between clinical and validity scales
substantial multicultural populations with individuals who
retain their traditional cultural orientations even as
5
biculturalism becomes more the norm in the immediate fu- Nunnally and Bernstein (1994) and Anastasi and Urbina (1997)
ture (Birman, 1994), along with increasingly large numbers have described procedures to identify differential validity of the
of genuinely international persons who possess skills and meaning of tests scores between groups. The relation between a pre-
dictor variable and a conceptually relevant criterion variable can be
knowledge from multiple cultural perspectives. In this way, explored for systematic differences between groups in the slope of
development of acculturation status norms would also help the regression line produced by this relationship. In cross-cultural
assure the clinical utility of the CS internationally while ad- and ethnic minority research, a significant difference in the correla-
vancing its multicultural clinical utility in the United States. tion coefficients between groups for a CS variable with a criterion
The Multicultural Assessment-Intervention Process model variable across the range of the CS variable scores would provide ev-
idence for a difference in the accuracy of its prediction of the crite-
(Dana, 2000b) procedure provides one model for their use in rion variable scores. This is termed slope bias and can be considered
clinical practice. evidence of differential validity for the variable across groups. How-
ever, a predictor variable can systematically overpredict or
underpredict the criterion variable for an ethnic minority group with-
out displaying a between-group difference in the degree of associa-
tion between the variables. Nunnally and Bernstein (1994) and
4
Acculturation refers to changes in the traditional culture patterns Lautenschlager and Mendoza (1986) have described two proce-
of groups resulting from continuous, first-hand contact, whereas ac- dures, moderated regression and step-down hierarchical regression,
culturation status refers to changes in the individual during this pro- for identifying both slope bias and this second type of systematic
cess as described by assimilation, biculturalism, marginalization, measurement error across groups, termed intercept bias. Intercept
and traditional cultural orientations (Dana, 1993). Acculturation sta- bias is an important and less often studied aspect of predictor bias;
tus roughly indicates the locus of an individual in a personal process Lopez (1989) noted the importance of study of both overprediction
of cultural change using orthogonal measurements of the extent to and underprediction of psychopathology, which has received less at-
which original culture characteristics are retained and new or host tention, across groups in cross-cultural and multicultural assessment
culture characteristics are incorporated (Cuéllar, 2000). research.
CROSS-CULTURAL RORSCHACH METHODOLOGY 197
when scores of African Americans were compared to underprediction of psychopathology reflects differential va-
Euro-Americans and, in the first study of both slope and in- lidity in the instruments tested, or if the instruments are instead
tercept bias with the MMPI–2 using the procedures of identifying bias among clinicians in their assignment of diag-
Lautenschlager and Mendoza (1986) in 65 comparisons, noses (in this case, overpathologizing behavior in ethnic mi-
found important Gender × Ethnicity interactions. Intercept nority individuals), or if the instruments have identified cases
bias was found in 32 MMPI–2 scale-criterion comparisons in which the diagnostic criteria are nonequivalent, or also
with men and 12 comparisons with women.6 Overprediction likely, that there exists some interaction of these factors.
of psychopathology was found in 7 cases, the remaining Several additional methodological considerations are rel-
cases involved underprediction. evant to the interpretation of findings from predictor bias
Meyer (2002) described a number of methodological limi- studies. Meyer (2002) noted that the cases of differential va-
tations to his study. Most important, Meyer (2002) noted his lidity he observed in his study (as in the case on all predictor
interpretation of findings (as with any predictor bias study) bias studies) could be related to one or more unmeasured and
assume the external correlates of behavior are equivalent. uncontrolled third variable (or variables; Anastasi & Urbina,
Meyer (2002) included education level, psychotic and de- 1977; Linn & Werts, 1971). This is a key point; we argue un-
pressive disorder diagnoses, maximum impairment associ- measured differences in variables associated with racial and
ated with the diagnosis, and composite MMPI–2 scores as cultural identity processes could comprise just such an im-
criterion variables; Arbisi et al. (2002) used comprehensive portant third variable in these predictor bias studies.7 All the
record review ratings with the additional methodological existing predictor bias studies in the literature provide lim-
strength of checks for interrater reliability, of demographics, ited description of their samples with regard to accultura-
Diagnostic and Statistical Manual of Mental Disorders tion/identity status; this represents an important empirical
(American Psychiatric Association, 1987 [DSM–III], 1994 question for future predictor bias studies. Nunnally and
[DSM–IV]) diagnosis, and 87 mood, affect, cognition, and Bernstein (1994) described how an analysis of the residuals
organic indicator ratings derived from the charting of the from the regression analysis can shed important light on po-
psychiatrist’s interview. The creation of criterion variables in tential third variables and assist with their identification.
these studies was well conceived and methodologically quite None of the existing predictor bias studies have reported re-
sophisticated; however, the variables rely on clinician re- sidual analysis, which would be a useful analytic strategy for
cords, and MMPI–2 scores as their data sources, all of which future studies. Further complicating the interpretation of re-
are the topic of ongoing controversy regarding cross-cultural gression analyses in predictor bias studies (and analysis of
equivalence. their residuals), Cuéllar (2000) discussed how culture can
For example, the cross-cultural accuracy of clinical judg- function as both a mediator variable influencing both the pre-
ment and diagnosis by clinicians has been a topic of immense dictor and criterion and a moderator variable altering the
historical controversy (Dana, 2001). Similarly, diagnosis and strength and direction of the relationship between predictor
psychopathology can differ in important ways and criterion. Because of the numerous complexities inher-
cross-culturally. In one example, Manson (1995) reviewed ent in the interpretation of predictor bias studies, although
factor analytic research with the Center for Epidemiological they provide important screening tests regarding the
Studies–Depression Scale (Radloff, 1977), a standardized, cross-cultural functioning of scales and measures, they inevi-
widely used composite measure of depressive tably leave unanswered important questions. These unre-
symptomatology developed by researchers at the National In- solved questions are best addressed through direct tests of the
stitute of Mental Health for epidemiological studies that in-
cludes items from several previously established depression
scales. Important differences emerged in the internal structure
of the instrument across Chinese American, Mexican Ameri- 7
Typically, controlling for a third variable is done to reduce vari-
can, and American Indian groups, which Manson interpreted ance extraneous to the criterion variable, which is assumed unrelated
as reflective of a change in the underlying patterning of affect to the predictor. Reducing extraneous variance allows a clearer test
in depression cross-culturally; this renders important facets of of the relationship between criterion and predictor. However, we are
the construct nonequivalent. In addition, regression can not including in our consideration of unmeasured third variable(s) those
establish causality, leaving unanswered several intriguing personality processes that underlie ethnic and cultural group mem-
bership, such as acculturation and cultural identity processes, or
questions. In the case where intercept bias indicates better still, specific psychological variables that underlie identity. In
underprediction, it remains unclear whether the observed this case, we are proposing third variables that both are related to
group status and are hypothesized to measure processes underlying
group status. This is an important point because simply treating an
identity variable as a third variable to be statistically controlled for
6
However, these differences, at least in part, were likely related to (e.g., partialed out) would reduce effects in regression analyses and
the sample size for women in the study (Arbisi et al., 2002) that was disguise its impact as a moderator variable when instead, covariance
comparatively smaller than the male sample; therefore, a larger ef- with predictor and criterion may be indication that it functions as the
fect size was required to attain significance among women. variable of interest underlying group status.
198 ALLEN AND DANA
cross-cultural construct validity and equivalence of the inter- a set of complex measures of personality structure and dynam-
pretations of data from test variables. ics including (a) attention, perception, and cognition; (b) the
experience and expression of emotion; and (c) attitudes toward
Multicultural and Cross-Cultural Construct self and others. A substantial body of literature from cultural
Equivalence and Construct Validation Research psychology has now conclusively demonstrated important
With the Rorschach ways in which normative personality attributes associated
with cognition, emotion, and self-representation vary widely
As the science of clinical psychology developed and empha- across cultural groups in intensity, form, and meaning (Church
sized an empirical basis for practice, the Rorschach remained & Lonner, 1998; Kitayama & Markus, 1994; Markus &
viable as an assessment instrument through efforts directed to- Kitayama, 1991; Mesquita & Frijda, 1992; Oyserman, Coon,
ward its adaptation into a test.8 However, Weiner (1994) as- & Kemmelmeier, 2002; Triandis & Suh, 2002). Some CS vari-
serted the Rorschach is a method rather than a test. According ables that comprise one component of the RIM likely tap cul-
to this view, the Rorschach inkblot method (RIM) is composed ture specific or emic constructs, other variables probably tap
of a variety of measures, scales, and indexes that tap numerous culturally universal or true etic constructs, and still other vari-
and often overlapping constructs, each supported by varying ables may include both emic and etic content (e.g., Populars).
degrees of validity data and each possessing different CS variables as well as the underlying constructs they tap may
psychometric properties, operating characteristics, and vary- be universal or culture specific. However, the cultural univer-
ing degrees of conformity to the assumptions of classical sality of specific Rorschach variables remains an empirical
psychometric theory. Therefore, the RIM is best understood as question.
This is not to suggest that research will fail to provide sup-
port for certain CS variables functioning as measures of etic
8
A test, as defined by Anastasi and Urbina (1997), is an objective constructs. For example, the process of introver-
measure of a sample of behavior standardized to ensure uniformity
sion–extraversion may represent a culturally universal or etic
of procedures using norms from a large, representative sample with
demonstrated reliability and validity predicated on independent ex- construct, perhaps even possessing similar distributions
ternal criteria. Dahlstrom (1993) provided additional specificity by across groups (Eysenck, 1991). Although Rorschach and
defining a psychological test to include the presence of six criteria: Exner (1993) both highlighted important ways in which
“(a) standardized materials and procedures, (b) optimal motivation, Erlebnistypus (EB) differs from the broader construct of in-
(c) immediate recording, (d) objective scoring, (e) appropriate
troversion–extraversion, it is reasonable to hypothesize that
norms, and (f) established validity” (p. 395), although we again must
clarify the important distinction that validity is not a property of a the two constructs share linking nomologicals or that EB
test but instead the interpretations made from test data, which can may even be a facet of the broader construct. Initial research
change based on the use intended. The CS makes significant inroads is encouraging regarding the potential for cultural universal-
toward fulfilling both the Anastasi and Urbina (1997) and the ity of the underlying dimensions tapped by EB (Dana, 1993;
Dahlstrom (1993) criteria for a test; however, numerous CS vari-
Dana et al., 1977), but further research is needed.
ables display violations of assumptions for the application of many
of the commonly used methods of inferential statistics used to test Weiner (2000) proposed that effective clinical applica-
their validity (e.g., numerous CS variables display non-normal dis- tion of the Rorschach as a method makes use of all relevant
tributions of scores that are skewed and kurtotic; cf. Viglione, 1995, Rorschach data in the formulation of interpretive state-
1997). In addition, issues fundamental to the Rorschach task system- ments to describe personality processes. Weiner (YEAR)
atically influence scores, most prominently the confounding influ-
emphasized a critical second stage in clinical interpretation
ence of individual differences in response frequencies (Meyer,
1992; Meyer & Archer, 2001). There exists limited detailed exami- with the RIM involving the specification of the levels of
nation of violation of assumptions such as normality and certainty associated with each interpretative statement. This
homoscedasticity for many of the statistical methods most widely requires strong assessor background and knowledge base in
used in validation work with most tests, scales, or items, including the empirical support for the specific Rorschach variables
those found on the Rorschach and MMPI–2. For example, recent
utilized in each interpretative statement. In contemporary
work in confirmatory factor analysis and linear structural modeling
provides evidence that many item and scale score distributions on multicultural Rorschach clinical application, an implication
numerous tests (or more precisely, their error distributions because of this viewpoint is that RIM interpretations are best devel-
the other components are typically constants) are exceedingly oped using nomothetic data along with idiographic data
skewed, or peaked/flattened, or both. However, we are also learning grounded in the cultural context of the assessee. Use of
that robust estimation procedures, particularly maximum likelihood,
nomothetic data requires a knowledge base of the
provide a reasonable remedy for this problem (West, Finch, &
Curran, 1995). In summary, “these [psychometric] methods require cross-cultural clinical utility of the CS interpretative vari-
only a definable population of numbers that meets the assumptions ables in question. Use of idiographic data is only possible
in the particular statistical method” (Nunnally & Bernstein, 1994, p. when the assessor is skilled and culturally competent in
23), and we are discovering many of our psychological measures in- work within the assessee’s cultural group. For assessment
cluding the MMPI–2 share with the Rorschach both numerous viola-
science in international and multicultural settings within the
tions of these assumptions and the existence of remedies for them.
Careful examination of these issues for most all scales and items is immediate future, nomothetic data sources from objective
still in its early stages. tests in conjunction with nomothetic and idiographic data
CROSS-CULTURAL RORSCHACH METHODOLOGY 199
sources from methods such as the Rorschach provide a Regression of Measures of Cultural Identity
framework for assessment practice. Such an approach holds on CS Variables
potential to combine instruments possessing culturally uni-
versal (etic) and culture specific (emic) elements guided by Cultural identity can be understood as a broad concept that
empirical findings from a common cross-cultural/multicul- subsumes the construct of acculturation and comprises an as-
tural research rubric (Dana, 2003). pect of self-concept (Rosenberg, 1979). Cultural identity en-
compasses the totality of thoughts and feelings that relate to
the social and ethnic identities associated with the cultural self.
METHODOLOGIES FOR CROSS- Broadly stated, measures of cultural identity tap important
CULTURAL/MULTICULTURAL VALIDATION facets of these social and ethnic identities. Many of the CS
RESEARCH WITH THE RORSCHACH variables are intended to tap dimensions of psychopathology.
If a CS variable tapping psychopathology, when regressed on a
Between-group comparisons of scores on Rorschach vari- measure of normative culture identity for members of that cul-
ables provide only a screening test of limited utility in investi- tural group, produces a statistically significant regression
gations of the hypothesized cultural universality of Rorschach equation, it suggests one of two possibilities: (a) cultural iden-
variables. Within the United States, accompanying accultura- tity measures are tapping identity development processes as-
tion data is necessary for these comparisons to have meaning. sociated with psychopathology, or (b) the variable is instead
To treat ethnicity as a demographic variable compromises the pathologizing cultural difference; in this case, normative cul-
internal validity of research designs, as the psychological vari- tural identity processes and culturally sanctioned modes of be-
able of interest is not measured but assumed to function havior explain a substantial portion of the variance in what the
through its demographic proxy variable. Cultural identity may CS variable is labeling as psychopathology. This becomes an
represent a more relevant variable when considering the use of important research question. Measures of psychopathology,
Rorschach normative data (Presley et al., 2001), and of even when conceptually equivalent within an ethnic minority
greater interest to personality researchers in both international group, generally should not covary with measures of norma-
cross-cultural and multicultural research are the specific per- tive group identity processes within the group (Allen & Walsh,
sonality processes underlying cultural identity that are respon- 2000).9 Should such association occur, given the potential for
sible for any observed differences on CS variables. Continued harm inherent in alternative (b), further research is needed to
development of this comparative basis would best be advanced identify which competing interpretation is accurate, and this
through national norms on such identity variables for coun- becomes an important additional component of the
tries in which significant differences exist on CS variables cross-cultural validation process for the variable. Routine ad-
compared to the U.S. norms, as with acculturation/cultural ministration of cultural identity variable measures in ethnic
identity status norms for the major multicultural groups within minority research would allow for routine tests of this impor-
the United States. tant question. By providing an understanding of the relation of
Currently, international CS norms are being complied in test variables to specific facets of identity processes and their
the absence of accompanying equivalence research. We associated psychological processes, the routine use of identity
have described a number of important methodological con- measures would move cross-cultural and ethnic minority as-
cerns associated with this absence of equivalence research sessment research in the direction of providing greater speci-
in the literature. Most prominent, a finding of difference in ficity in the variables of interest, which are the psychological
norms can mean several things including the measure or its variables that underlie group membership for which group
underlying construct has changed or the groups differ on membership as a demographic variable serves as proxy
the construct of interest. Similarly, the instance in which (Okazaki & Sue, 1995). For CS variables that do not tap
CS normative data matches across cultural groups is not a psychopathology, such a relationship with a cultural identity
direct empirical test of cross-cultural construct equivalence measure could either represent cultural difference or a shift in
and is prone to the effects of numerous confounding vari- the nomological network underlying the construct across cul-
ables. Predictor bias testing, although a useful methodolog- tural groups; in this latter case, the variable is not
ical advance on comparative studies that focus on cross-culturally transportable.
differences between scores across groups, has its own set of
methodological difficulties, most prominent, the assump-
tion that the criterion variable used is equivalent. In light of
these important methodological considerations that impact 9
This is not meant to assert that it can never be the case that the
the interpretability of even well designed comparative stud- prevalence of psychopathology varies between racial, ethnic, or cul-
ies, research is also needed that screens specific CS vari- tural groups or in the more restricted sense that psychopathology can
never vary with normative cultural identity within a group. Both
ables in a more rigorous manner for potential cross-cultural
cases are possible, however, as we outline in this article, numerous
construct nonequivalence and then directly tests the associated methodological issues and empirical questions must be
cross-cultural construct validity of these CS variables. addressed before such conclusions can be drawn.
200 ALLEN AND DANA
As is the case with MMPI–2 research, most Rorschach of similarity among principal components loadings adjusted
studies to date with ethnic minorities treat ethnicity as a de- for family-wise error rate. Because the combination of numer-
mographic variable. However, the variable of interest in psy- ous different groups with limited description of their cultural
chology is instead the specific culturally embedded and ethnic backgrounds into one non-Euro-American group
psychological variables responsible for an observed differ- potentially obscures important differences between different
ence on a Rorschach score. An important goal of a ethnic minority groups, replication of these promising find-
cross-cultural and multicultural Rorschach research program ings with separate samples from each of the distinct ethnic mi-
to guide practice is an understanding of the role of specific nority groups in the United States is needed. This type of
culturally mediated variables that underlie identity and ac- analysis is also termed the assessment of measurement
count for observed group differences in CS variables that tap invariance across groups. Jöreskog and Sörbom (1989) de-
cross-culturally equivalent constructs. For example, facets of scribed extensions of CFA for use with multiple samples well
the culturally mediated variable of self-concept, such as indi- suited to the study of measurement invariance. Multigroup
vidualism–collectivism (Oyserman et al., 2002), may poten- CFA is currently possible using existing international data de-
tially contribute to an understanding of meaning of the scribed by Meyer (2001) and T. Schaeffer (personal commu-
observed cultural difference in scores on a particular CS vari- nication, November 3, 2002) and would provide for an
able and in addition, an enhanced cross-cultural understand- immediate and important test of the factorial invariance of the
ing of the underlying construct tapped by the variable. Both CS across multiple international samples.
of these understandings are goals of cross-cultural and ethnic Thissen, Steinberg, and Wainer (1988) provided a brief dis-
minority assessment research as articulated in the method- cussion of IRT’s uses in the study of DIF across groups and in-
ological perspectives offered by Okazaki and Sue (1995). cluded brief discussion of direct tests of hypotheses regarding
item parameters. Although widely utilized in intelligence test-
Factor Analytic and Item Response Theory ing, IRT can be equally effectively used in personality testing.
Approaches However, IRT has only been sporadically applied in research
on personality instruments (Panter, Swygert, Dahlstrom, &
Much of the controversy in cross-cultural and multicultural Tanaka, 1997); our review of the literature could locate no IRT
use of the Rorschach and other psychological tests has fo- study with the CS. In the only published IRT study with the
cused on evidence or lack of evidence of bias. In multicul- MMPI–2, Childs, Dahlstrom, and Panter (2000) discovered
tural and cross-cultural personality assessment, one way to DIF on the MMPI–2 by gender in the measurement of depres-
conceptualize many of the commonly described types of bias sion using Scale 2; this same methodology used to test gender
(e.g., Van de Vijver & Poortinga, 1997) is through the logic differences by Childs et al. is identically applicable in tests of
of construct validation. Test bias, from this viewpoint, occurs cultural differences. IRT holds great promise for study of the
when a test does not measure equivalent constructs across differential functioning for several CS variables because
groups; this can be the result of the underlying construct many of the CS variables are not true scales as articulated by
shifting between cultural groups or the measure in question the tenets of classical psychometric theory (e.g., Nunnally &
displaying different measurement characteristics across Bernstein, 1994). IRT provides an alternative to classical mea-
groups, leading interpretations from test data to possess dif- surement theory, providing an avenue to more rigorous study
ferential validity across groups. Viewed from this light, the of CS variables across cultural and ethnic groups. One draw-
central question in multicultural and cross-cultural research back to the use of IRT has been the requirement for large sam-
with the Rorschach or any other standard psychological as- ple sizes. However, Meijer (2003) discussed the use of
sessment instrument is that of cross-cultural construct valid- person-fit statistics that allow one to study the fit of an individ-
ity. Tests of CS structural validity, most notable confirmatory ual item pattern to an IRT model. If a significant proportion of
factor analysis (CFA), and tests of differential item function- a small cultural or ethnic sample were identified as misfitting
ing (DIF), most notable item response theory (IRT) analysis the IRT model produced by the normative sample for a CS
described elsewhere (Allen & Walsh, 2000) along with the variable, it would suggest a pattern of person misfit to the
regression of cultural identity measures on CS variables can model that is associated with group status and in this way pro-
provide screening tests to identify promising CS variables for vide a test of item equivalence cross-culturally (see van der
the more painstaking process of cross-cultural validation. Flier, 1982). However, Allen and Walsh (2000) described four
Meyer (2002) provided an initial structural test of the CS important limitations to the use of IRT procedures in what is
with a sample of 432 Euro-American, African American, His- commonly termed the investigation of DIF across ethnic and
panic American, Asian American, and American Indian psy- cultural groups.10
chiatric patients matched on key demographic variables. No
significant differences emerged when the factor structure of 10
A number of important limitations exists regarding IRT proce-
67 key CS variables produced by Euro-Americans was com-
dures in tests of item equivalence. Two important assumptions must
pared to that of a composite sample of non-Euro-Americans be met for the use of IRT with personality instruments. First, the set
using Jensen’s (1980) method of computation of coefficients of items to be analyzed is assumed to be unidimensional as demon-
CROSS-CULTURAL RORSCHACH METHODOLOGY 201
Cross-Cultural Construct Validation 1995, 2001a, 2001b). In particular, key issues to be attended
to include interrater reliability (Meyer, 1997a, 1997b; Meyer
Because of limitations to each of the previously described et al., 2002; Viglione & Hilsenroth, 2001; Weiner, 1991) and
methods, the most unequivocal method to establish absence the use of appropriate statistical methods (Mcguire, Kindler,
of bias for the inferences from a CS variable cross-culturally Curtiss, & Viglione, 1995; Viglione, 1995).
is to establish evidence of cross-cultural construct equiva-
lence of the variable with the cultural group of interest Mixed Quantitative/Qualitative-Ethnographic
through compelling validity argument. Predictor bias stud- Research Designs With the Rorschach
ies, tests of structural equivalence, and IRT analyses all pro-
vide useful screens to identify promising variables to devote In their pilot study of the role of clitoridectomy in structuring
energies toward the more painstaking task of cross-cultural the self-concept of West-African immigrant women in Paris,
construct validation. It is quite reasonable to assume that France from the Soninké culture, Franchi and
norms for some CS variables may differ cross-culturally, and Andronikof-Sanglade (1999) used a combined emic–etic ap-
this normative difference may reflect only cultural difference proach that provides one promising model for innovative
in quantity of an attribute within an equivalent construct. The cross-cultural validity research with the Rorschach. In
critical question is does the variable measure a Franchi and Andronikof-Sanglade’s study, self-report mea-
cross-culturally equivalent construct? Returning to the exam- sures of variables associated with their hypotheses were ad-
ple of EB, what is important here is evidence for the construct ministered along with the Rorschach. Data from the quantita-
equivalence of the underlying introversive–extratensive Ror- tive CS measures was examined contextually through the
schach dimension, not whether scores on EB differ idiographic narrative data produced by the Rorschach proto-
cross-culturally. Such research could include cross-cultural col. In one example from Franchi and
replication of earlier validation studies with EB and its com- Andronikof-Sanglade’s approach, alternative inferences re-
ponents, M and Color responses (e.g., Blatt & Feirstein, garding the positive SCZI scores produced by a majority of
1977; Cocking, Dana, & Dana, 1969; Dana, 1968; Dana & the participants (SCZI was the precursor to the current PTI in
Cocking, 1968; Dana, Cocking, & Dana, 1970; Exner & the CS) were developed out of the narrative data and accom-
Sanglade, 1992; Weiner & Exner, 1991; and see Exner, 1993; panying quantitative measures. With no other indication of
Molish, 1967; and Singer, 1960 for a review of earlier re- psychopathology in their nonpatient Soninké sample,
search literature). Methodological recommendations from Franchi and Andronikof-Sanglade inferred a shift had oc-
the mainstream Rorschach research literature are essential in curred in the nomological net underlying the construct
guiding cross-cultural studies (e.g. Exner, 1995; Weiner, tapped by SCZI within their sample and accordingly devel-
oped new inferences based in the narratives, the self-report
variables, and interestingly, other Rorschach variables tap-
strated empirically through factor analysis. CS variables intended to
measure a particular construct are often probably multidimensional, ping cognitive functioning. Further research with
and this is often the case with MMPI–2 scales. Whereas Drasgow West-African women is necessary, and not all of the Franchi
and Parsons (1983) suggested IRT is relatively robust to violations and Andronikof-Sanglade’s speculations may receive sup-
of unidimensionality, others (e.g., Panter et al., 1997) have asserted port in future studies. Nonetheless, Franchi and
that IRT is best used with empirically identified factor scales to en- Andronikof-Sanglade’s study highlights the utility of the
sure properties of the IRT model are maintained in the analysis. Sec-
ond, germane to the MMPI–2, local independence must hold for the multiple quantitative and qualitative data sources produced
item set, that is, item responses must depend on or be based on their by a Rorschach protocol in cross-cultural research.
common association with θ, the underlying personality trait or at- Many authors (e.g., Brink, 1994; Kleinman, 1988) have
tribute, rather than other external variables related to the items them- argued that the exclusive reliance on quantitative approaches
selves such as the potential effects of item order, content carryover can result in research that misses or potentially misunder-
from other items, or use of common stem. Another limitation of IRT
involves the relatively large sample sizes needed for stable parame- stands important depth and nuance in investigations of cul-
ter estimation, a major disadvantage for researchers who study small tural factors. Methodologies that integrate qualitative
overall population groups. Because of this, adaptation of person-fit approaches with nomothetic assessment research are sorely
statistics to detect item pattern score misfit among a proportion of needed because of their utility in discovery-based research
cultural group members holds promise. Finally, IRT yields findings and initial validation work with psychological constructs.
regarding item equivalence and not conceptual equivalence of the
construct. This is an important distinction. Item equivalence on a test Qualitative methods in more mature and advanced areas of
indicates that individual items function similarly across cultures or cultural research can provide important triangulation and
ethnic groups and that the observed score metrics are equivalent. checks on quantitative findings as well as potentially enrich
Item equivalence does not imply construct or conceptual equiva- and extend them. In the Franchi and Andronikof-Sanglade
lence. IRT does not provide a test of whether it may be appropriate to (1999) study, qualitative-idiographic data provided a critical
adopt the nomological network of the construct of interest to the new
cultural or ethnic group. Further, the previously mentioned applica- check that prevented erroneous interpretations from the
tions of IRT cannot provide information regarding the source of quantitative data and raised important questions regarding
DIF. the construct equivalence of particular variables. Franchi and
202 ALLEN AND DANA
Andronikof-Sanglade’s study demonstrates both the value of Consequential Aspects of Cross-Cultural
qualitative method design components in cross-cultural per- and Multicultural Rorschach Validity Research
sonality assessment research along with the utility of the
Rorschach in cultural research through its applicability to During three decades of work, Messick (1965, 1975, 1980,
mixed quantitative–qualitative designs. 1995) developed a comprehensive approach to construct vali-
dation in which he has described six aspects of construct valid-
ity. Our discussion of cross-cultural research methodologies
CURRENT STATUS OF THE CROSS- with the CS has thus far been primarily occupied with only five
CULTURAL/MULTICULTURAL aspects: content validity through discussion of strategies to in-
RORSCHACH RESEARCH vestigate the cross-cultural CS stimulus materials and instruc-
tional set relevance; substantive validity through proposed
study of whether the CS task and variables actually engage
In their review, Costantino, Flanagan, and Malgady (1995)
processes tapped by the construct cross-culturally; structural
described numerous methodological deficiencies in current
validity through cross-cultural investigation of the invariance
cross-cultural and multicultural Rorschach research. At that
of the internal structure of CS variables; generalizability,
time, most research was at the stage of cross-ethnic and
which includes study both of the generalizability of CS inter-
cross-cultural comparisons of variable scores. Costantino et
pretive frames across cultural groups and their cross-cultural
al. noted the circularity of using the variables themselves as
clinical utility; and external validity, which encompasses pro-
criteria in the absence of demonstrated universality of the un-
cedures to establish cross-cultural and multicultural conver-
derlying constructs. Recent research with the CS described
gent and discriminant validity of specific CS variables.
previously has begun to move beyond these problematic
Messick (1995) highlighted a final aspect of validity, the
comparative methodologies, and we present methodologies
consequential aspect, which elevates a consideration of val-
here to advance this research further.
ues and meaning as integral to the test validation process.
Use of the Rorschach, as with any test or measure, should be
Multicultural and Cross-Cultural Clinical appraised in light of the potential and actual social conse-
Utility Research quences of test interpretation consistent with other social val-
ues. Important consequential validity considerations arise in
McGrath (2001) differentiated between clinical utility and cross-cultural interpretations of test data. Throughout the
validity research with the Rorschach. Many of McGrath’s past century of psychological research, cultural differences
distinctions are useful and relevant to cross-cultural and have been typically interpreted as deficiencies (Jones, 1988).
multicultural research with the Rorschach. Indeed, much of Because of this, consequential validity considerations point
the debate regarding the merits of the CS focuses on clini- to the need for development of an empirical basis for multi-
cal utility. McGrath discussed the issue of categorical pre- cultural and cross-cultural practice grounded in consensual
diction and positive and negative predictive power in rela- research standards. Standards currently exist for U.S. minori-
tion to the accuracy of an instrument’s prediction of clinical ties (CNPAAEMI, 2000) and provide a model for interna-
categories such as diagnostic categories. Returning yet tional cross-cultural research.
again to PTI as example, in cross-cultural research, the re-
search question from the perspective of clinical utility re-
search translates to “once we have established PTI taps an CONCLUSIONS
equivalent construct with a new cultural group, does PTI
display equivalent predictive power for the diagnosis of Consequential validity elevates investigations of multicultural
schizophrenia within a new cultural group?” However, ex- and cross-cultural clinical utility as a key component of con-
perience with SCZI among the Soninké highlights the im- temporary assessment instrument validation. The Rorschach
portance of first establishing conceptual equivalence of a enjoys a long history of cross-cultural application, both as a
variable before proceeding to this next stage of analysis of clinical instrument and measure of cultural difference, begin-
cross-cultural predictive precision. ning with Herman Rorschach’s early research interests. The
Moreover, although McGrath (2001) described clinical nonverbal nature of its stimulus materials uniquely situates it
utility as a distinct consideration from conventional validity as a true international assessment instrument of underutilized
theory and research, an alternative perspective is to view potential. However, development of an empirical basis for
clinical utility as an important and critical component of va- cross-cultural and multicultural Rorschach practice is needed
lidity theory as originally proposed by its primary architects. to realize this promise. We have reviewed here selected meth-
Cook and Campbell (1979) described the generalizability as- odological considerations for such efforts. Additionally,
pect of construct validity, which concerns itself directly with cross-cultural and multicultural Rorschach research can offer
its utility within the various settings a measure is used. This an important contribution in the contemporary debate regard-
generalizability aspect includes the settings and demands of ing the clinical utility of the CS. The cross-cultural and multi-
actual clinical use of an instrument. cultural potential of the Rorschach provides an important ra-
CROSS-CULTURAL RORSCHACH METHODOLOGY 203
tionale for its continued use in the 21st century. Equally Butcher, J. N., Dahlstrom, W. G., Graham, J. R., Tellegen, A., & Kaemmer, B.
(1989). MMPI–2: Minnesota Multiphasic Personality Inventory–2: Manual
important, continued work with the CS can also provide mod-
for administration and scoring. Minneapolis: University of Minnesota Press.
els for the development of new, performance-based, interna- Butcher, J. N., Nezami, E., & Exner, J. (1998). Psychological assessment of
tional personality measures with nonverbal materials de- people in diverse cultures. In S. S. Kazarian & D. R. Evans (Eds.), Cul-
signed for cross-cultural transportability. tural clinical psychology: Theory, research, and practice (pp. 61–105).
New York: Oxford University Press.
Butcher, J. N., & Rouse, S. V. (1996). Personality: Individual differences and
clinical assessment. Annual Review of Psychology, 45, 87–111.
ACKNOWLEDGMENTS Camara, W., Nathan, J., & Puente, A. (2000). Psychological test usage: Im-
plications in professional use. Professional Psychology: Research and
This research has been supported in part by a grant from the Practice, 31, 141–154.
National Institute of Alcohol Abuse and Alcoholism and the Childs, R. A., Dahlstrom, W. G., & Panter, A. T. (2000). Item response the-
ory in personality assessment: A demonstration using the MMPI–2 De-
National Center for Minority Health Disparities
pression Scale. Assessment, 7, 37.
1RO1AA11446–03 (principal investigator: G. Mohatt; Church, A. T., & Lonner, W. J. (1998). The cross-cultural perspective in the
coinvestigators: K. Hazel, J. Allen, & C. Geist). We thank study of personality: Rationale and current research. Journal of
Mark J. Hilsenroth, James A. Walsh, and four anonymous re- Cross-Cultural Psychology, 29, 32–62.
viewers for their helpful comments on earlier versions of this Cocking, R., Dana, J., & Dana, R. H. (1969). Six constructs to define Ror-
schach M: A response. Journal of Projective Techniques & Personality
manuscript.
Assessment, 33, 322–323.
Cook, T. D., & Campbell, D. T. (1979). Quasi-experimentation: Design and
analysis issues for field settings. Chicago: Rand McNally.
REFERENCES Costantino, G., Flanagan, R., & Malgady, R. (1995). The history of the Ror-
schach: Overcoming bias in multicultural projective assessment.
Allen, J., & Walsh, J. A. (2000). A construct-based approach to equivalence: Rorschachiana, 20, 148–171.
Methodologies for cross-cultural/multicultural personality assessment re- Council of National Psychological Associations for the Advancement of Eth-
search. In R. H. Dana (Ed.), Handbook of cross-cultural and multicultural nic Minority Interests. (2000). Guidelines for research in ethnic minority
personality assessment (pp. 63–85). Mahwah, NJ: Lawrence Erlbaum As- communities. Washington, DC: American Psychological Association.
sociates, Inc. Cronbach, L. J. (1949). Statistical methods applied to Rorschach scores: A
American Psychiatric Association. (1980). Diagnostic and statistical man- review. Psychological Bulletin, 46, 393–429.
ual of mental disorders (3rd ed.). Washington, DC: Author. Cronbach, L. J. (1971). Test validation. In R. L. Thorndike (Ed.), Educa-
American Psychiatric Association. (1987). Diagnostic and statistical man- tional measurement (2nd ed., pp. 433–507). Washington, DC: American
ual of mental disorders (3rd ed., rev.). Washington, DC: Author. Council on Education.
American Psychiatric Association. (1994). Diagnostic and statistical man- Cuéllar, I. (2000). Acculturation as a moderator of personality and psycho-
ual of mental disorders (4th ed.). Washington, DC: Author. logical assessment. In R. H. Dana (Ed.), Handbook of cross-cultural and
American Psychological Association. (2002). Ethical principles of psychol- multicultural personality assessment (pp. 113–129). Mahwah, NJ: Law-
ogists and code of conduct. American Psychologist,57, 1060–1073. rence Erlbaum Associates, Inc.
Anastasi, A., & Urbina, S. (1997). Psychological testing (7th ed.). Upper Dahlstrom, W. G. (1993). Tests: Small samples, large consequences. Ameri-
Saddle River, NJ: Prentice Hall. can Psychologist, 48, 393–399.
Arbisi, P., Ben-Porath, Y., & McNulty, J. (2002). A comparison of MMPI–2 Dana, R. H. (1955). Rorschach scorer reliability. Journal of Clinical Psy-
validity in African American and Caucasian psychiatric inpatients. Psy- chology, 11, 401–403.
chological Assessment, 14, 3–15. Dana, R. H. (1962). The validation of projective tests. Journal of Projective
Arnold, B. R., Montgomery, G. T., Castanada, I., & Longoria, R. (1994). Ac- Techniques, 26, 182–186.
culturation and performance of Hispanics on selected Halstead–Reitan Dana, R. H. (1968). Six constructs to define Rorschach M. Journal of Pro-
neuropsychological tests. Assessment, 1, 239–248. jective Techniques and Personality Assessment, 32, 138–145.
Atkinson, D. R. (1983). Ethnic similarity in counseling psychology: A re- Dana, R. H. (1993). Multicultural assessment perspectives for professional
view of research. The Counseling Psychologist, 11, 79–92. psychology. Boston: Allyn & Bacon.
Azibo, D. A. (1988). Understanding the proper and improper usage of the Dana, R. H. (2000a). Culture and methodology in personality assessment. In
comparative framework. Journal of Black Psychology, 15, 81–91. I. Cuéllar & F. Paniagua (Eds.), Handbook of multicultural mental health:
Beutler, L. E., Crago, M., & Arizmendi, T. G. (1986). Therapist variables in Assessment and treatment of diverse populations (pp. 97–120). San
psychotherapy process and emotions. In A. E. Bergin & S. L. Garfield Diego, CA: Academic.
(Eds.), Handbook of psychotherapy and behavior changes (pp. 257–310). Dana, R. H. (Ed.). (2000b). Handbook of cross-cultural and multicultural
New York: Wiley. personality assessment. Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Birman, D. (1994). Acculturation and human diversity in a multicultural soci- Dana, R. H. (2001). Clinical diagnosis of multicultural populations in the
ety. In E. J. Trickett, R. J. Watts, & D. Birman (Eds.), Human diversity: Per- United States. In L. Suzuki, J. Ponterotto, & P. Meller (Eds.), The hand-
spectives on people in context (pp. 261–284). San Francisco: Jossey-Bass. book of multicultural assessment (2nd ed., pp. 101–131). San Francisco:
Blatt, S. J., & Feirstein, A. (1977). Cardiac response and personality organi- Jossey-Bass.
zation. Journal of Consulting and Clinical Psychology, 45, 111–123. Dana, R. H. (2003). Assessment training, practice, and research for the new
Bornstein, R. F. (2001). Clinical utility of the Rorschach inkblot method: millennium: Challenges and opportunities for professional psychology.
Reframing the debate. Journal of Personality Assessment, 77, 39–47. Ethical Human Sciences and Services, 5, 127–140.
Brink, T. L. (1994). The need for qualitative research on mental health of el- Dana, R. H., Aragon, M., & Kramer, T. (2002). Public sector mental health
derly Hispanics. International Journal of Aging and Human Develop- services for multicultural populations: Bridging the gap from research to
ment, 38, 279–291. clinical practice. In M. N. Smyth (Ed.), Health care in transition (Vol. 1,
Butcher, J. N. (Ed.). (1996). International adaptations of the MMPI–2: Re- pp. 15–29). Hauppauge, NY: Nova Science.
search and clinical applications. Minneapolis: University of Minnesota Dana, R. H., & Cocking, R. (1968). Cue parameters, cue probabilities, and
Press. clinical judgment. Journal of Clinical Psychology, 24, 475–480.
204 ALLEN AND DANA
Dana, R. H., Cocking, R., & Dana, J. (1970). The effects of experience and Hall, G. C. N., Bansal, A., & Lopez, I. R. (1999). Ethnicity and
training on accuracy and configural analysis. Journal of Clinical Psychol- psychopathology: A meta-analytic review of 31 years of comparative
ogy, 26, 28–32. MMPI/MMPI–2 research. Psychological Assessment, 11, 186–197.
Dana, R. H., Hinman, S., & Bolton, B. (1977). Dimensions of examinee re- Hibbard, S. (2003). A critique of Lilienfeld et al.’s (2000) “The scientific sta-
sponses to the Rorschach: An empirical analysis. Psychological Reports, tus of projective techniques.” Journal of Personality Assessment, 80,
40, 1147–1153. 260–271.
Dawes, R. M. (1999). Two methods for studying the incremental validity of Hilsenroth, M. J., & Handler, L. (1995). A survey of graduate students’ expe-
a Rorschach variable. Psychological Assessment, 11, 297–302. riences, interests, and attitudes about learning the Rorschach. Journal of
Draguns, J. G. (1990). Applications of cross-cultural psychology in the field Personality Assessment, 64, 243–257.
of mental health. In R. W. Brislin (Ed.), Applied cross-cultural psychology Hoffmann, T., Dana, R. H., & Bolton, B. (1985). Measured acculturation and
(pp. 302–324). Newbury Park, CA: Sage. MMPI–168 performance of Native American adults. Journal of
Drasgow, F., & Parsons, C. K. (1983). Application of unidimensional item Cross-Cultural Psychology, 16, 243–256.
response models to multidimensional data. Applied Psychological Mea- Howard, A., Poin, G. M., Sechrest, L. B., Cordray, D. S., Kaplan, L., Hall, J.,
surement, 7, 189–199. et al. (1987). Membership opinions about reorganizing APA. American
Exner, J. E. (1993). The Rorschach: A Comprehensive System: Vol. I. Basic Psychologist, 42, 763–779.
foundations (3rd ed.). New York: Wiley. Hunsley, J., & Bailey, J. M. (1999). The clinical utility of the Rorschach: Un-
Exner, J. E. (Ed.). (1995). Issues and methods in Rorschach research. fulfilled promises and an uncertain future. Psychological Assessment, 11,
Mahwah, NJ: Lawrence Erlbaum Associates, Inc. 266–277.
Exner J. E., & Sanglade, A. A. (1992). Rorschach changes following Jensen, A. R. (1965). Review of the Rorschach Inkblot Test. In O. K. Buros
brief and short term therapy. Journal of Personality Assessment, 59, (Ed.), The sixth mental measurements yearbook (pp. 501–509). Highland
59–71. Park, NJ: Gryphon.
Eysenck, H. J. (1959). Review of the Rorschach Inkblot Test. In O. K. Buros Jensen, A. R. (1980). Bias in mental testing. New York: Free Press.
(Ed.), The fifth mental measurements yearbook (pp. 276–278). Highland Jones, J. M. (1988). Racism in black and white: A bicultural model of reac-
Park, NJ: Gryphon. tion and evolution. In P. A. Katz & D. A. Taylor (Eds.), Eliminating rac-
Eysenck, H. J. (1991). Dimensions of personality: The biosocial approach to ism: Profiles in controversy (pp. 117–135). New York: Plenum.
personality. In J. Strelau & A. Angleitner (Eds.), Explorations in tempera- Jöreskog, K. G., & Sörbom, D. (1989). LISREL 7: User’s reference guide.
ment: International perspectives on theory and measurement (pp. Mooresville, IN: Scientific Software.
87–103). New York: Plenum. Kazarian, S. S., & Evans, D. B. (1998). Cultural clinical psychology. In S. S.
Flier, H. van der (1982). Deviant response patterns and comparability of test Kazarian & D. R. Evans (Eds.), Cultural clinical psychology: Theory, re-
scores. Journal of Cross-Cultural Psychology, 13, 267–298. search, and practice (pp. 1–38). New York: Oxford University Press.
Franchi, V., & Andronikof-Sanglade, A. (1999). Methodological and Kitayama, S., & Markus, H. R. (1994). Emotions and culture: Empirical
epistemological issues raised by the use of the Rorschach comprehensive studies of mutual influence. Washington, DC: American Psychological
system in cross-cultural research. Rorschachiana, 23, 118–134. Association.
Frueh, B. C., Gold, P. B., de Arellano, M. A., & Brady, K. L. (1997). A racial Kleinman, A. (1988). Rethinking psychiatry: From cultural category to per-
comparison of combat veterans evaluated for PTSD. Journal of Personal- sonal experience. New York: Free Press.
ity Assessment, 68, 692–702. Lautenschlager, G. J., & Mendoza, J. L. (1986). A step-down hierarchical
Frueh, B. C., Smith, D. W., & Libet, J. M. (1996). Racial differences on psy- multiple regression analysis for examining hypotheses about test bias in
chological measures in combat veterans seeking treatment for PTSD. prediction. Applied Psychological Measurement, 10, 133–139.
Journal of Personality Assessment, 66, 41–53. Lilienfeld, S. O., Wood, J. M., & Garb, H. N. (2000). The scientific status of
Gacono, C. B., Loving, J. L., & Bodholdt, R. H. (2001). The Rorschach and projective techniques. Psychological Science in the Public Interest, 1,
psychopathology: Toward a more accurate understanding of the Ror- 27–66.
schach findings. Journal of Personality Assessment, 77, 16–38. Lindsey, M. L. (1998). Culturally competent assessment of African Ameri-
Gamst, G., Dana, R. H., Der-Karabetian, A., & Kramer, T. (2000). Ethnic can clients. Journal of Personality Assessment, 70, 43–53.
match and client ethnicity effects on global assessment and visitation. Lindzey, G. (1954). Projective techniques and cross-cultural research. New
Journal of Community Psychology, 28, 547–564. York: Appleton-Century-Crofts.
Gamst, G., Dana, R. H., Der-Karabatian, A., & Kramer, T. (2001). Asian Linn, R. L., & Werts, C. E. (1971). Considerations for studies of test bias.
American mental health clients: Effects of ethnic match and age on global Journal of Educational Measurement, 8, 1–4.
assessment and visitation. Journal of Mental Health Counseling, 23, Lonner, W. J. (1996). Psychometrics and culture. In Advanced methodological
57–71. issues in culturally competent evaluation for substance abuse prevention.
Ganellen, R. J. (2001). Weighting evidence for the Rorschach’s validity: A Washington, DC: SAMHSA, Center for Substance Abuse Prevention.
response to Wood et al. (1999). Journal of Personality Assessment, 77, Lopez, S. R. (1989). Patient variable biases in clinical judgment: Conceptual
1–15. overview and methodological considerations. Psychological Bulletin,
Garb, H. N., Wood, J. W., Lilienfeld, S. O., & Nezworski, M. T. (2002). Ef- 106, 184–203.
fective use of projective techniques in clinical practice: Let the data help Maki, M. T. (1999). The effects on clinician identification when clinician
with selection and interpretation. Professional Psychology, 33, and client share a common ethnic minority background. Journal of Multi-
454–463. cultural Social Work, 7, 57–72
Garb, H. N., Wood, J. W., Nezworski, M. T., Grove, W. M., & Stejskal, W. J. Malgady, R. G. (1996). The question of cultural bias in assessment and diag-
(2001). Toward a resolution of the Rorschach controversy. Psychological nosis of ethnic minority clients: Let’s reject the null hypothesis. Profes-
Assessment, 13, 433–448. sional Psychology: Research and Practice, 27, 73–77.
Geisinger, K. F. (1994). Cross-cultural normative assessment: Translation Manson, S. M. (1995). Culture and major depression: Current challenges in
and adaptation issues influencing the normative interpretation of assess- the diagnosis of mood disorders. Psychiatric Clinics of North America,
ment instruments. Psychological Assessment, 6, 304–312. 18, 487–501.
Graham, J. R. (1990). MMPI–2: Assessing personality and Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for
psychopathology. London: Oxford University Press. cognition, emotion, and motivation. Psychological Review, 98, 224–253.
Greene, R. L. (1987). Ethnicity and MMPI performance: A review. Journal McFarlane, J. W. (1942). Problems in validation inherent in projective meth-
of Consulting & Clinical Psychology, 55, 497–512. ods. American Journal of Orthopsychiatry, 12, 405–410.
CROSS-CULTURAL RORSCHACH METHODOLOGY 205
McGrath, R. E. (2001). Toward more clinically relevant assessment re- Handbook of cross-cultural and multicultural personality assessment (pp.
search. Journal of Personality Assessment, 77, 122–127. 247–266). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Mcguire, H., Kindler, B. N., Curtiss, G., & Viglione, D. J. (1995). Some spe- Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.).
cial issues in data analysis. In J. Exner (Ed.), Issues and methods in Ror- New York: McGraw-Hill.
schach research (pp. 227–250). Mahwah, NJ: Lawrence Erlbaum Associ- Okazaki S., & Sue, S. (1995). Methodological issues in assessment research
ates, Inc. with ethnic minorities. Psychological Assessment, 7, 367–375.
McNulty, J. L., Graham, J. R., Ben-Porath, Y., & Stein, L. A. R. (1997). Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking indi-
Comparative validity of MMPI–2 scores of African American and Cauca- vidualism and collectivism: Evaluation of theoretical assumptions and
sian mental health center clients. Psychological Assessment, 9, 464–470. meta-analyses. Psychological Bulletin, 128, 3–72.
Meijer, R. R. (2003). Diagnosing item score patterns on a test using item re- Panter, A. T., Swygert, K. A., Dahlstrom, W. G., & Tanaka, J. S. (1997). Fac-
sponse theory-based person-fit statistics. Psychological Methods, 8, tor analytic approaches to personality item-level data. In J. A. Schinka &
72–87. R. L. Greene (Eds.), Emerging issues and methods in personality assess-
Mensh, I. N. (1950). Statistical techniques in present day psychodiagnostics. ment (pp. 285–307). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Psychological Bulletin, 47, 474–492. Piotrowski, C. (1999), Assessment practices in the era of managed care:
Mesquita, B., & Frijda, N. H. (1992). Cultural variations in emotions: A re- Current status and future directions. Journal of Clinical Psychology, 55,
view. Psychological Bulletin, 112, 179–204. 787–796.
Messick, S. (1965). Personality measurement and the ethics of assessment. Pires, A. A. (2000). National norms for the Rorschach normative study in
American Psychologist, 20, 136–142. Portugal. In R. H. Dana (Ed.), Handbook of cross-cultural and multicul-
Messick, S. (1975). The standard problem: Meaning and values in measure- tural personality assessment (pp. 367–392). Mahwah, NJ: Lawrence
ment and education. American Psychologist, 30, 955–966. Erlbaum Associates, Inc.
Messick, S. (1980). Test validity and the ethics of assessment. American Pollack, D., & Shore, J. H. (1980). Validity of the MMPI with Native Ameri-
Psychologist, 35, 1012–1027. cans. American Journal of Psychiatry, 137, 946–950.
Messick, S. (1995). Validity of psychological assessment: Validation of in- Pope-Davis, D. B., Liu, W. M., Toporek, R. L., & Brittan-Powell, C. S.
ferences from persons’ responses and performances as scientific inquiry (2001). What’s missing from multicultural competency research: Review,
into score meaning. American Psychologist, 50, 741–749. introspection, and recommendations. Cultural Diversity and Ethnic Mi-
Meyer, G. J. (1992). Response frequency problems in the Rorschach: Clini- nority Psychology, 7, 121–138.
cal and research implications with suggestions for the future. Journal of Pritchard, D. A., & Rosenblatt, A. (1980). Racial bias in the MMPI: A meth-
Personality Assessment, 58, 231–244. odological review. Journal of Consulting and Clinical Psychology, 48,
Meyer, G. J. (1997a). Assessing reliability: Critical corrections for a critical 263–267.
examination of the Rorschach Comprehensive System. Psychological As- Presley, G., Smith, C., Hilsenroth, M., & Exner, J. (2001). Rorschach valid-
sessment, 9, 480–489. ity with African Americans. Journal of Personality Assessment, 77,
Meyer, G. J. (1997b). Thinking clearly about reliability: More critical cor- 491–507.
rections for a critical examination of the Rorschach Comprehensive Sys- Radloff, L. S. (1977). A CES–D scale: A self-report scale for research in the
tem. Psychological Assessment, 9, 495–498. general population. Applied Psychological Measurement, 1, 385–401.
Meyer, G. J. (2001). Evidence to correct misperceptions about Rorschach Ritsher, J., Slivko-Kolchik, E., & Oleichik, I. (2001). Assessing depression
norms. Clinical Psychology: Science and Practice, 8, 389–396. in Russian psychiatric patients: Validity of MMPI and Rorschach. Assess-
Meyer, G. J. (2002). Exploring possible ethnic differences and bias in the ment, 8, 373–389.
Rorschach Comprehensive System. Journal of Personality Assessment, Ritzler, B., & Alter, B. (1986). Rorschach teaching in APA-approved clinical
78, 104–129. graduate programs: Ten years later. Journal of Personality Assessment,
Meyer, G. J., & Archer, R. P. (2001). The hard science of Rorschach re- 50, 44–49.
search: What do we know and where do we go? Psychological Assess- Rorschach, H. (1942). Psychodiagnostics: A diagnostic test based on per-
ment, 13, 486–502. ception (3rd ed., B. Kronenberg & P. Lemkau, Trans.) Berne, Switzerland:
Meyer, J. G., Finn, S. E., Eyde, L. D., Kay, G. G., Moreland, K. L., Dies, R. Hans Huber. (Original work published in 1921)
R., et al. (2001). Psychological testing and psychological assessment: A Rosenberg, M. (1979). Conceiving the self. New York: Basic.
recent review of evidence and issues. American Psychologist, 56, Roysircar-Sodowsky, G., & Kuo, P. Y. (2001). Determining cultural validity
128–165. of personality assessment. In D. B. Pope-Davis & H. L. K. Coleman
Meyer, J. G., Hilsenroth, M. J., Baxter, G., Exner, J. E., Jr., Fowler, C. J., (Eds.), The intersection of race, class, and gender in multicultural coun-
Piers, C. C., et al. (2002). An examination of interrater reliability for scor- seling (pp. 213–239). Thousand Oaks, CA: Sage.
ing the Rorschach Comprehensive System in eight data sets. Journal of Sangro, F. L. (1997). Location tables, form quality, and popular responses in
Personality Assessment, 78, 219–274. a Spanish sample of 470 sujects. Rorschachiana, 22, 38–66.
Miller, J. B., Rosenthal, R., Bornstein, R. F., Berry, D. T. R., & Sanz, I. A. E., Ritzler, B. A., Ivanouw, J., & Dana, R. H. (2001). El impacto
Brunell-Neuleib, S. (1999). A comparative meta-analysis of Rorschach de las neuvas technologies en el futuro de las technicas de evaluacion
and MMPI validity. Psychological Assessment, 11, 278–296. psicologica clinica [The impact of new technologies on the future of clini-
Molish, H. B. (1967). Critique and problems of the Rorschach. A survey. In cal psychological assessment]. Revista Argentina de Clinica Psicologica,
S. J. Beck & H. B. Molish (Eds.), Rorschach’s test. II: A variety of person- X, 203–225.
ality pictures (2nd ed.). New York: Grune & Stratton. Singer, J. L. (1960). The experience type: Some behavioral correlates and
Montgomery, G. T., & Orozco, S. (1985). Mexican Americans’ performance theoretical implications. In M. Rickers-Ovsiankina (Ed.), Rorschach psy-
on the MMPI as a function of level of acculturation. Journal of Personal- chology (pp. 325–372). Huntington, NY: Kreiger.
ity Assessment, 54, 328–342. Stout, C. E. (1997). Psychological assessment in managed care. New York:
Muniz, J., Prieto, G., Almeida, L., & Bartram, D. (1999). Test use in Spain, Wiley.
Portugal, and Latin American countries. European Journal of Psychologi- Stricker, G., & Gold, J. R. (1999). The Rorschach: Toward a nomothetically
cal Assessment, 15, 151–157. based, idiographically applicable configurational model. Psychological
Munroe, R. L. (1945). Objective methods and the Rorschach blots. Ror- Assessment, 11, 240–250.
schach Research Exchange, 9, 59–73. Sue, S., Keefe, K., Enomoto, K., Durvasula, R. S., & Chao, R. (1996). Asian
Nichols, D. S., Padilla, J., & Gomez-Maqueo, E. L. (2000). Issues in the American and White college students’ performance on the MMPI–2. In J.
cross-cultural adaptation and use of the MMPI–2. In R. H. Dana (Ed.), N. Butcher (Ed.), International adaptations of the MMPI–2: Research
206 ALLEN AND DANA
and clinical applications (pp. 206–218). Minneapolis: University of Min- Weiner, I. B., Spielberger, C. D., & Abeles, N. (2002). Scientific psychology
nesota Press. and the Rorschach Inkblot Method. The Clinical Psychologist, 55(4), 7–12.
Sue, S., Kurasaki, K. S., & Srinivarsan, S. (1999). Ethnicity, gender, and West, S. G., Finch, J. F., & Curran, P. J. (1995). Structural equation models
cross-cultural issues in clinical research. In P. C. Kendall, J. N. Butcher, & with nonnormal variables: Problems and remedies. In R. H. Hoyle (Ed.),
G. N. Holinbeck (Eds.), Handbook of research methods in clinical psy- Structural equation modeling: Concepts, issues, and applications (pp.
chology (2nd ed., pp. 54–71). New York: Wiley. 56–75). Thousand Oaks, CA: Sage.
Sue, S., & Zane, N. (1987). The role of culture and cultural techniques in Whatley, J. P., Allen, J., & Dana, R. H. (2003). Racial identity and the MMPI
psychotherapy: A critique and reformulation. American Psychologist, 42, in African American male college student. Cultural Diversity and Ethnic
37–45. Minority Psychology, 9, 344–352.
Thissen, D., Steinberg, L., & Wainer, H. (1988). Use of item response theory Whitworth, R. H., & McBlaine, D. D. (1993). Comparison of the MMPI and
in the study of group differences in trace lines. In H. Wainer & H. I. Braun MMPI–2 administered to Anglo-American and Hispanic American uni-
(Eds.), Test validity (pp. 213–238). Hillsdale, NJ: Lawrence Erlbaum As- versity students. Journal of Personality Assessment, 61, 19–27.
sociates, Inc. Widiger, T. A. (2001). The best and worst of us? Clinical Psychology: Sci-
Timbrook, R. E., & Graham, J. R. (1994). Ethnic differences on the MMPI? ence and Practice, 8, 374–377.
Psychological Assessment, 6, 212–217. Wood, J. M., Garb, H. S., Lilienfeld, S. O., & Nezworski, M. T. (2002). Clin-
Triandis, H. C., & Suh, E. M. (2002). Cultural influences on personality. An- ical assessment. Annual Review of Psychology, 53, 519–543.
nual Review of Psychology, 53, 133–160. Wood, J. M., & Lilienfeld, S. O. (1999). The Rorschach Inkblot Test: A case
Van de Vijver, F., & Hambleton, R. K. (1996). Translating tests: Some prac- of overstatement? Assessment, 6, 341–351.
tical guidelines. European Psychologist, 1, 89–99. Wood, J. M., Lilienfeld, S. O., Garb, H. N., & Nezworski, M. T. (2000a).
Van de Vijver, F. J. R., & Poortinga, Y. H. (1997). Towards an integrated Limitations of the Rorschach as diagnostic tool: A reply to Garfield
analysis of bias in cross-cultural assessment. European Journal of Psy- (2000), Lerner (2000), and Wiener (2000). Journal of Clinical Psychol-
chological Assessment, 13, 29–37. ogy, 56, 441–448.
Viglione, D. J. (1995). Basic considerations regarding data analysis. In J. Wood, J. M., Lilienfeld, S. O., Garb, H. N., & Nezworski, M. T. (2000b). The
Exner (Ed.), Issues and methods in Rorschach research (pp. 227–250). Rorschach test in clinical diagnosis: A critical review with a backward look
Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. at Garfield (1947). Journal of Clinical Psychology, 56, 395–430.
Viglione, D. J. (1997). Problems in Rorschach research and what to do about Wood, J. M., Lilienfeld, S. O., Nezworski, M. T., & Garb, H. N. (2001).
them. Journal of Personality Assessment, 68, 590–599. Coming to grips with the negative evidence for the Comprehensive sys-
Viglione, D. J. (1999). A review of recent research addressing the utility of tem for the Rorschach: A comment on Gacono, Loving, and Bodholdt;
the Rorschach. Psychological Assessment, 11, 251–265. Ganellen; and Bornstein. Journal of Personality Assessment, 77, 48–70.
Viglione, D. J., & Hilsenroth, M. J. (2001). The Rorschach: Facts, fictions, Wood, J. M., Nezworski, M. T., Garb, H. N., & Lilienfeld, S. O. (2001a). The
and future. Psychological Assessment, 13, 452–471. misperception of psychopathology: Problems with the norms of the Com-
Vinet, E. V. (2000). The Rorschach Comprehensive System in Iberoamerica. prehensive System for the Rorschach. Clinical Psychology: Science and
In R. H. Dana (Ed.), Handbook of cross-cultural and multicultural per- Practice, 8, 350–373.
sonality assessment (pp. 345–365). Mahwah, NJ: Lawrence Erlbaum As- Wood, J. M., Nezworski, M. T., Garb, H. N., & Lilienfeld, S. O. (2001b).
sociates, Inc. Problems with the norms of the Comprehensive System for the Ror-
Voigt, W., & Dana, R. H. (1964). Inter- and intrascorer reliability. Journal of schach: Methodological and coneptual consideration. Clinical Psychol-
Projective Techniques and Personality Assessment, 28, 92–95. ogy: Science and Practice, 8, 397–402.
Wainer, H., & Braun, H. I. (Eds.). (1988). Test validity. Hillsdale, NJ: Law- Wood, J. M., Nezworski, M. T., & Stejskal, W. J. (1996). The Comprehen-
rence Erlbaum Associates, Inc. sive System for the Rorschach: A critical examination. Psychological Sci-
Watkins, C. E., Jr., Campbell, V. I., Nieberding, R., & Hallmark, R. (1995). ence, 7, 3–10.
Contemporary practice of psychological assessment by clinical psycholo- Wood, J. M., Nezworski, M. T., Stejskal, W. J., & Garven, S. (2001). Ad-
gists. Professional Psychology: Research and Practice, 26, 54–60. vancing scientific discourse in the controversy surrounding the Compre-
Wechsler, D. (1997). Manual for the Wechsler Adult Intelligence Scale III. hensive Systems for the Rorschach: A rejoinder to Meyer (2000). Journal
New York: The Psychological Corporation. of Personality Assessment, 76, 369–378.
Weiner, I. B. (1991). Editor’s note: Interscorer agreement in Rorschach re- Wood, J. M., Nezworski, M. T., Stejskal, W. J., Garven, S., & West, S. G.
search. Journal of Personality Assessment, 56, 1. (1999). Methodological issues in evaluating Rorschach validity: A com-
Weiner, I. B. (1994). The Rorschach Inkblot Method (RIM) is not a test: Im- ment on Burns and Viglione (1996), Weiner (1996), and Ganellen (1996).
plications for theory and practice. Journal of Personality Assessment, 62, Assessment, 6, 115–129.
498–504. Wuts, G. (2002). A preliminary validation study of the psychometric proper-
Weiner, I. B. (1995). Methodological considerations in Rorschach research. ties of selected scales of the MMPI–2 with English speaking
Psychological Assessment, 7, 330–337. Singaporeans. Unpublished doctoral dissertation, California School of
Weiner, I. B. (1999). Contemporary perspectives on Rorschach assessment. Professional Psychology, Alliant University, San Diego.
European Journal of Personality Assessment, 15, 78–86.
Weiner, I. B. (2000). Making Rorschach interpretation as good as it can be.
James Allen
Journal of Personality Assessment, 74, 164–174.
Weiner, I. B. (2001a). Advancing the science of psychological assessment: Department of Psychology
The inkblot method as exemplar. Psychological Assessment, 13, University of Alaska Fairbanks
423–432. Fairbanks, AK 99775–6480
Weiner, I. B. (2001b). Considerations in collecting Rorschach reference E-mail: Jim.Allen@uaf.edu
data. Journal of Personality Assessment, 77, 122–127.
Weiner I. B., & Exner, J. E. (1991). Rorschach changes in long-term and
sort-term psychotherapy. Journal of Personality Assessment, 56, Received April 23, 2003
453–465. Revised September 26, 2003

View publication stats

You might also like