You are on page 1of 10

JOURNAL OF PERSONALITY ASSESSMENT, 81(1), 64–73

Copyright © 2003, Lawrence Erlbaum Associates, Inc.

Modifying the Rorschach Human Experience Variable
to Create the Human Representational Variable VIGLIONE, PERRY,
RORSCHACH
JANSAK,
HRV
MEYER, EXNER

Donald J. Viglione
California School of Professional Psychology, San Diego
Alliant International University

William Perry
Department of Psychiatry
University of California, San Diego

Donna Jansak
California School of Professional Psychology, San Diego
Alliant International University

Gregory Meyer
Department of Psychology
University of Alaska Anchorage

John E. Exner, Jr.
Rorschach Workshops
Asheville, North Carolina

The Human Experience Variable (HEV; Perry & Viglione, 1991) provided information about
interpersonal perceptions not previously available within the Comprehensive System (CS;
Exner, 1993). Research data suggests that it was related to interpersonal functioning and as a re-
sult, psychological impairment and health. In this article, we present the rationale and empirical
basis for recent psychometric refinements to the HEV, consequently renamed the Human Rep-
resentational Variable (HRV). Research addressing the reliability and validity for the HRV is
summarized. Based on data and experience with the HEV, this study summarized some small
modifications to the original algorithm. The refined variable, the HRV, has been added to the
CS (Exner, 2000). Data presented here suggest that the HRV has improved psychometric prop-
erties compared to the HEV and that it is simpler to understand. Research recommendations
and interpretive suggestions are also presented.

Object relational, attachment, and interpersonal schema the- human representational and interactional Rorschach re-
ories suggest that Rorschach human representations are sponses are related to interpersonal behavior and patterns in
linked through real-life perceptions to interpersonal related- both pathological and nonpathological populations (Blake,
ness as a major factor involved in psychological health or dis- Humphrey, & Feldman, 1994; Burns & Viglione, 1996;
turbance (Blatt, Brenneis, Schimek, & Glick, 1976; Blatt & Fowler, Hilsenroth, & Handler, 1996; Fritsch & Holmstrom,
Lerner, 1983; Mayman, 1967). Confirming this point of 1990; Hart & Hilton, 1988; Hibbard, Hilsenroth, Hibbard, &
view, a great amount of empirical data associates various as- Nash, 1995; Ryan, Avery, & Grolnick, 1985; Urist, 1977;
pects of Rorschach human representations with psychologi- Urist & Shill, 1982; for negative evidence, see Blatt &
cal health versus impairment (e.g., see summaries by Exner, Berman, 1990). Adapting this work to the Comprehensive
1993; Perry & Viglione, 1991; Stricker & Healy, 1990; System (CS; Exner, 1993), Picker (1984) incorporated CS
Viglione, 1999; Weiner, 1966). Research results suggest that form quality and special scores with Urist’s (1977) human

” Wo F– (Hd) 5. Thus. and logical features. and arms and stuff tential and to adapt the literature fully to the CS. in this article we present minor modifica. descriptions of human experience reflect the respondent’s lized existing CS variables that were associated with human understanding of people and relationships. PHR (or PHE with previous tions to the HEV to produce a new CS variable. with another group of nonpatient women with (1985. PHE responses incorporated distorted. The HEV itself was a weighted difference personal relations.0 COP. This finding is especially notable in that restrictions developed to represent the object representation aspect of of range within a nonpatient sample would lead one to expect a ego functioning (Perry & Viglione. confused. neurophysiological dysfunction. others. HEV = Human Experience Variable. (Burns. the CS did not have a summary variable that orga- nized the interpersonal perception information available on view. The left side of Table 2 presents the HEV algorithm. This the Rorschach to the quality of interpersonal relationships algorithm relied on CS response codes to identify human ex. Rorschach responses involving human content and the test. Perry & Viglione. researchers have as- ventional. To create such a score. benign or cooperative interaction (or GHE with previous coding) man perception and representation on the Rorschach. According to this small effect size in a nonpatient sample even if the study em- . tion with psychological health and impairment (Adrian & ignated as GHE when they were scored Popular. GHE = Good Human Experience. logical. Their PHE/PHR: Distorted. the Human coding) Representational Variable (HRV). damaged. damaged. 1996. The interrater In contrast. The response also be understood from other theoretical points of view. GHE responses incorporated accurate. HRV = Human Representational Variable. Perry presents conceptual examples of good and poor human re. Perry & Viglione. responses were classified as either (a) positive/intact.03) or when standardized scores or Z-Score transformations of PHE and computed using weights based on standardized scores derived GHE. psychological disturbance. from the distributions in the Burns and Viglione data (Cohen’s From a theoretical perspective. 1991). 1995. a monster or goal. 1991). or (b) nega. The weights relatively more negative or problematic interpersonal rela- and a constant were derived to equalize the contribution of the tionships. GHR = Good Human Representation. of people and relationships helps one to visualize and to de- rithm. relating the respondents’ human representations on The HEV was operationally defined by an algorithm.97. was to increase reliability imaginary. the middle is blown aggressive human response or away and you can see some body and precision of measurement so as to increase validity po- representation parts. Also. Poor Human Experience responses (PHE). Based on this GHE/GHR: Accurately perceived. malevo. response pull and com. Within the HEV algo. and illogical features. lent. fictional and partial human responses were des. Continuing spread around. aggressive. there is a strong empirical grounding for Rorschach Basis Example human representations as a measure of interpersonal percep- tion and psychological health or impairment. & Braff. Thus. This TABLE 1 research reported an association between Picker’s CS-based HEV and HRV Conceptual Basis and Examples score and interpersonal or object-relations development. relative to previous scales. The HEV could representational features of the responses. 1996). realistic. In the past. cognitive theory suggests that one’s understanding research and conceptual descriptions. and measures of mon responses were considered in developing the algorithm. con. reliability on the HEV. Hh P 3. Researchers found the HEV to be related to severity of positive or negative features. Card III: “Two brothers setting a realistic. Also. there was essentially no difference in the effect size when cal- Specifically. Table 1 Kaser-Boyd. scribe such phenomena in Rorschach test responses because Good Human Experience responses (GHE). sessed the reliability and validity of the HEV. GHE and PHE responses within the HEV summary variable. Card III: “Upside down.5 in this tradition. thus supporting the connec- For example. prognosis. The effect size for this comparison was large. Of course. the HEV. as described by their spouses and them- score developed from data published in Haller and Exner selves.01). Other research focused on the criterion-related validity of fied HRV. MOR. GHR the Human Experience Variable (HEV) as a measure of hu. PHR = Poor HEV Human Representation. Perry. & Braff. and interpersonal relations. Cadenhead. 1994. Note. Perry and Viglione (1991) uti. any given response did not incorporate all these 1991). the HEV in- tive/problematic. the HEV algorithm was a difference between the culated with the original weights (Cohen’s d =1. is excellent (intraclass correlation = . 1993. legs. logical. RORSCHACH HRV 65 representation scale. volved the schema of self. Since it was originally developed. very few cues are available from the test. or alien person. the Mutuality of Autonomy scale. For characteristics used in the HEV were derived from previous example. sponses as classified by either the original HEV or the modi. Burns and Viglione perience responses and to differentiate between GHE and contrasted 35 nonpatient women with strong and stable inter- PHE responses. Burns & Viglione. PHE = Poor Human Experience. benevolent. the HEV was originally d = 1. Haller. Conceptually. unrealistic. Perry and Viglione (1991) developed responses or representation with 1 Mao 2 H. 1991). Perry & Viglione. 1982. intact human table for a party at their house” D + complex of findings. when scored by well-trained coders. intact.

or (b) An Hd coding [not (Hd) coding] 7. Assign PHE to any remaining responses that have either 4. (H). FQo. Assign Good (GHE) for responses containing a Pure H coding that 1. 1999). ex- analysis with logistic regression addressed incremental valid.. Assign Poor (PHR) to the remaining responses that have either: (a) FQ minus. To and psychological health. . Express as ratio of GHE to PHE responses (e.66 VIGLIONE. or FQu (b) No cognitive special scores (b) No cognitive Special Scores except DV (c) No Special Scores AG or MOR 2. This analysis demonstrated that the HEV’s ability to dif.. Also. X – %. stant in the original HEV calculation equation. (Hd). Assign PHR to any remaining responses that have either: (a) The Special Scores AG. PHE was greater in number and more variable than GHE. CONTAM. Assign GHR to any remaining responses that have the Special Score (a) Popular to III. sonal perception as a foundation to interpersonal relatedness Nezworski. ment of the responses according to interpretive guidelines Stejskal. 5:3) 1. INCOM1. Assign GHR to all remaining responses Step C: Create the GHE to PHE Ratio and Calculate the HEV Step C: Create the GHR to PHR Ratio and Calculate the HRV 1. Hd. suggest that Rorschach human representations cap. INCOM.51(PHE) – 0. 1999). or (b) FABCOM1. As in the original (Perry & Viglione. HEV = Human Experience Variable. algorithm for differentiating GHE from PHE responses with TABLE 2 HEV and HRV Algorithms for Classifying Human Representational Responses As Good or Poor HEV HRV Step A. Stejskal. DR. IV. (a) The Special Scores of FABCOM or MOR. (a) FQ minus or FQ none (no Form) or (b) ALOG. 1991) study. Such a simplification could only be justified if the GHE and ranged from low to high HEV scores. or (b) ALOG. from healthy to PHE distributions were more similar. MEYER. these weights confused some commentators (Wood. or any Level 2 cognitive Special Score. Assign Good (GHR) for responses containing a Pure H coding that also have all of the following: also have all of the following: (a) Form Quality of FQ+. this predomi. problematic scores. Garven. PERRY. EXNER ployed an extreme groups design (Wood. Wood et al. or FQu (a) Form Quality of FQ+. it was determined that a sim- Table 3 summarizes the available HEV descriptive data ple raw (unit weighted) difference score would be preferable. PHE = Poor Human Experience. For these reasons. Assign GHE to any remaining responses that have either 3. Assign Either a Good or Poor to Each of These Human Responses Through the Following Classification Algorithm 1. Calculate the HEV score 2. Moreover. Select All Responses That Contain Either Human Content Coding [H. The groups are ar. IV. or COP or AG Special Scores With an FM Coding Step B. THE HRV For example. by other Rorschach variables and responses (WSum6. 5:3) 2.g. CONTAM. that is. Hd. or (b) The Content of An (c) Responses with FQnone 5. PHR = Poor Human Representation. and the responses excluded from the frequency of PHE relative to GHE initially led Perry and HEV).g. & West. or IX 6. However. make calculation more simple. Hx]. VII. neutral or even positive human representational responses ferentiate these two groups of women were not accounted for were misclassified as PHE. in the context of the previous positive Viglione (1991) to incorporate corrective weights and a con- findings. These findings. (Hd). tured by the HEV offered specific information about interper. Nezworski. or any Level 2 cognitive Special Score (c) AG or MOR 3. aminations of representative protocols suggested that some ity. Express as ratio of GHR to PHR responses (e. GHR = Good Human Representation. The greater variability of PHE associated with the higher demographic variables. 2001. FQo. we decided to examine the original HEV nance of PHE in many cases seemed to contradict our assess. & Garven. Post hoc goodness-of-fit culled from the empirical and theoretical literature. JANSAK. Assign Poor (PHE) to the remaining responses that have either 2. Assign PHR to any remaining responses that have either: (a) Responses without H that contain (H). VII.75(GHE) + 0.. Assign GHE to all remaining responses 5. These research findings and experience with the ENHANCING THE HEV TO PRODUCE HEV revealed some possible areas for improving the HEV. DR1. Calculate the HRV score (a) Formula 0.04 = HEV (a) Formula GHR – PHR = HRV (b) Positive scores are associated with more impairment (b) Negative scores are associated with more impairment Note. PHE responses were more frequent than GHE responses in almost all samples. published in journals and dissertations. Assign GHR to any remaining responses that have a Popular to III. GHE = Good Human Experience. or Hx. the Determinant M. and IX or COP but do not have the Special Score AG (b) Responses with COP 4. HRV = Human Representational Variable.

68 1. and (Hd) ing research projects with nonpatients controls and for indi- viduals with depression. et al.85 3.94 Viglione.48 –0.43 Ephebophile Roman Catholic priests 79 5. PHR = Poor Human Representation..13 1. and if possible.11 2.27 0. Sprock. dSamples from the Adrian & Kaser-Boyd (1995) study share participants..04 Perry. GHR = Good Human Representation.12 3.90 –0. major 22 2. 1994 locked residential-care ward Adrian & Kaser-Boyd.00 3.33 3. RORSCHACH HRV 67 TABLE 3 HEV and HRV Scores of Adults Published in Journal Articles or Dissertations GHE PHE HEV Source Sample Description N M SD M SD M SD HEV Burns.37 3.40 Perry.92 2. problematic interpersonal relationships (individuals with sponses relative to poor human responses.23 1. therefore. with late-life onset.35 4. or poor. terpersonal relationships (nonpatients/controls).60 3.40 0.45 amphetamines DeLucas. mean age = 75 45 2. 1995d Inpatients 61 1.48 3. 1989.84 3.97 0. We examined Human Detail. aEight participants in this dissertation were included in the sample (N = 389) described later in this article.10 1. In Table 4 these groups are ordered from human representational responses that are accompanied by . 2000 Older nonpatients. Available tify a simple difference score.86 –0.01 1.58 Netter. Netter & Inpatient. Burns & Nonpatient married women poor 35 2.40 1. 1997 Military violent male offenders 33 2. the aim of (a) increasing the frequency of good human re.35 1. from the factor analysis. 1990.35 0.50 2.65 2.35 1. 1995 Nonpatient college men placebo 20 2.83 0. and Fictional Human available Rorschach data in computerized forms from ongo. We reviewed some readily available data to provide direction for changes to the HEV algorithm.00 1. nonpsychotic 48 2.23 adequate and good relationships Auslander. Sprock.32 3. 1994 Nonpatient volunteers 20 3.80 2. et al.10 0. 1991 Haller. eThe HRV is scored in the opposite direction of the HEV.15 2.90 stabilized on meds. Table arranged from lowest to highest HEV scores and highest to lowest HRV scores.91b Netter. PHE = Poor Human Experience.63 1. 1995d Outpatients 24 2. Netter & Viglione. 1993. (H).00 2. GHE = Good Human Experience.04 1.23 2.30 2.20 2. or The percentage scores in Table 4 refer to the percentage of schizophrenia. 1996 interpersonal relations Adrian & Kaser-Boyd.60 5.30 1.90 2.59 –0.09 1.03 1. 1997a Military male security personnel— 30 2.80 1.66 2. 1996 interpersonal relations DeLucas.39 2. 1990. bHRV estimated from GHE and PHE.60 0.38 1. (c) increasing demographic information for these groups is presented in the the validity of the classification of human responses as good footnotes of Table 4. Contents. M = 0 and SD = 1. 1997 Military nonviolent male offenders 32 2. small dose 20 2.39 Adrian & Kaser-Boyd.29b depression GHR PHR HRVe M SD M SD M SD HRV McGlone. Burns & Nonpatient married women good 35 3. 1995d Clinical team diagnosis. 1985c Inpatients with depression symptoms 50 2. High HRV scores and low HEV scores are associated with more positive interpersonal perception.30 2.06 2. 1995d Clinical team diagnosis.40 4.78 1. 1993.16 2.20 3.40 3.20 2.00 2.09 Note.00 1.98b Viglione. HRV = Human Representational Variable.93 about to deploy overseas DeLucas. mean age = 60 Perry.85 3. Haller & Exner.02 1.00b Adrian & Kaser-Boyd. schizophrenia diagnosis from 20 1.40 2. 2001 Control Roman Catholic priests 80 4. 44 2.10 4. HEV = Human Experience Variable.26 0. 1993 Women whose military husbands were 68 2.75 –1. Hd.21 1.59 Viglione.80 3. 1995 Nonpatient college men.58 3.63 1.30 1. 1995d Clinical team diagnosis. 1982.02b Ingham.44 1.15 1.85 1.42 1.25 0.60 1.00c 1. cThis is the HEV derivation sample.86 3.43 2.00 –0.36b Burns. a history of criminal offenses.20 1.86 Pedophile Roman Catholic priests 78 4.70 1.28 1.37 1.32b Auslander. (b) making the schizophrenia) to more positive and mutually enhancing in- good and poor human distributions more similar so as to jus.79 2.20 3. psychotic 37 1.98 Viglione.50 4.80 1.12 –0. 2000 Older patients with schizophrenia. Perry & Outpatients with depression diagnosis 49 2.10 2.00c Adrian & Kaser-Boyd.27 1.

age M = 36. In other words. In the sample of Step B. (1996). 3 unknown. As a secondary consequence of this move. e40 males and 53 females. (Hd). Based on the findings in Table 4 with these various Level 1 Cognitive Special Scores human content categories. Morgan & Viglione (1992). Percentages correspond to percentage of all Human Element relevant responses with the variable noted in the column heading. The differences between the groups the algorithm. the responses entailed positive or negative interpersonal ages at the bottom of the columns. only 2% of the human responses reduce the importance of AG and MOR. greater percent. Viglione. and Hx as poor criteria but retained Hd as a poor criterion for the new The Level 1 cognitive Special Scores of INCOM1. were classified as good based on having a COP on this step of we examined many representative responses with MOR and TABLE 4 Percentages of Human Responses With Hd Versus Other Human Contents H Hd (H) + (Hd) + Hx H – [Hd + (H) + (Hd) + Hx] H – Hd Groupa N GHE Step B. K. AG. DeLucas (1997). JANSAK. c55 males and 27 females. it also contributes to our sec. sponses with both COP and MOR are now classified as good and (H). EXNER these human contents. there was a basis to 294 nonpatient controls. 34% of all human representational responses among the schizophrenia group included a Pure H content. 4 Asian Americans. age M = 35. and INCOM1 nents. 29 females. GHR = Good Human Representation. (2001). MOR. in the new HRV. Strauss. (2003). with different interpersonal capacity. SD = 14. 3b. the algorithm. (Hd) in the HRV algorithm preserves its more valid compo. Rorschach protocols were derived from samples previously reported in Brinderson. HRV Step B. The second to last column in HEV algorithm. 1b PHE Step B. This expected pattern was not discernible for good based on the occurrence of a COP. & Gowri. 6a. Montemagni. For this group and others. (Hd). is discernible for the HRV schema. it was not often used in the algorithm to cial Scores were found early in the HEV algorithm (HEV classify good responses in HEV Step B. (AG) and Morbid (MOR) did not differ greatly. bSteps and algorithm refer to HEV steps in Table 2. Accordingly. ethnicity data not available. (1995). presumably with with (b) the percentage of human responses classified by healthier interpersonal relationships. (1997). PHE = Poor Human Experience. whereas they were assigned PHE in the ble 2. PHR = Poor Human Representation. ethnicity was 5 African Americans. To investigate this. age M = 39. These Spe- sonal relatedness. MEYER. Gaudiana. The right side of Table 2 provides the complete algo. Netter & Viglione (1994). To ensure this crite- seen in Step B. 6b and 7. Hd. 72 White. 4b to ond goal for the HRV of increasing the number of good re. re- In the HEV algorithm. DR1. HRV = Human Representational Variable. groups at the bottom of the table. 4a). GHE = Good Human Experience. Viglione. Aggressive Movement tively high frequency among samples with superior interper. Because responses are no longer classified as poor and DR1 scores were retained at the end of the algorithm. 4a PHE Poor Step B. we dropped (H). Pure H identified good responses. To classify more responses as and less Hd. ethnicity data not available. sentations in the FABCOM1 responses were more impaired taining Hd to identify poor responses while dropping (H) and then those in the INCOM1 and DR1 responses. the results for rion was working properly. we moved COP up in (H). sponses relative to poor. and Hx in the third column of the table. A. Perry.68 VIGLIONE. Table 5 presents (a) are small but in the expected direction for H and Hd. re. SD = 12. 2c) as poor criteria. Accordingly. produced more Pure H COP in the HEV algorithm. Table 2 this can be seen by comparing HEV Step B. f127 males and 167 females. FABCOM1 were grouped together in the last step of the rithm for scoring the HRV and the changes just noted can be HEV algorithm as poor human criteria. Table 4 represents this HEV algorithm for the four clinical samples. with 71 missing age data. 4a HEV Algorithmb HRV Algorithm Schizophreniac 85 34 24 28 –18 10 Offendersd 243 39 24 29 –14 15 Depressione 93 40 22 32 –14 18 Nonpatient/Controlsf 294 40 19 30 –19 21 Note. Viglione. PERRY. & Braff (1992). aThese groups are listed in supposed order of more intact human representational capacity from hypothetically most disturbed (schizophrenia) to most intact (nonpatient/controls). Our collective judgment was that the human repre- algorithm but not for the HEV algorithm. 1999. 8 Hispanic American. and Hx identified poor responses (see Ta. and HRV. Accordingly. that is. Other demographic information not available in computerized Rorschach files. HEV column. 4 unknown ethnicity. Step B. d214 males. In other the percentages of human responses with COP contrasted words. Jansak. Returning to Table 4. et al. In based on (H) and (Hd) scores. we examined numerous re- this component of the HRV algorithm are represented in the sponses that were classified at this step to determine whether final column. Brager. (Hd). and An COP Our data indicated that among various groups presumably Although Cooperative Movement (COP) occurs at a rela. SD = 11. FABCOM1 was moved up in the algorithm. HEV = Human Experience Variable. The expected pattern. .

1997) Approximately one fourth of this HRV devel- Method and Results opmental sample was included in the preliminary HRV anal- ysis presented earlier in this article. 363 contained more than 13 re- ble 6 contains the relevant descriptive and demographic data sponses and were retained in the final sample. Viglione. > 54 46 12 15 Missing data 1 1 Ethnicity AG. Montemagni. nonpatients 50 13 Community volunteers High scores for the HRV are in the positive or desirable di- mild-moderate depressed 40 10 rection. 1999. RORSCHACH HRV 69 TABLE 5 TABLE 6 Percentages Associated With COP in HEV Sample Descriptive Data (Step B.S. In other words. Within each subgroup. Ex. which subgroups. We conducted a study to compare the psychometric proper- ties of the original HEV with the HRV and to determine how closely the variables are related. for the entire subgroup. Gaudiana. Perry. which incorporates the minor enhancements CS character disorders 48 12 summarized previously. the right hand column in Table 2 presents the new Status CS nonpatients 50 13 HRV algorithm. The HRV consists of two variables. 3c in Table 2) 1997/1998 % of Human Responses Census COP As % of Classified As Good by COP N % Estimatea Group Human Responses in the HEV Sex Schizophrenia 5 1 Female 193 50 Offenders 7 1 Male 192 49 Depression 8 2 Missing data 4 1 Nonpatient/ Age (years) Controls 12 2 18 to 25 77 20 15 26 to 54 265 68 70 Note. 2003. we were unable to identify viduals were slightly underrepresented. As indicated Missing data 66 17 earlier. for these subgroups demo- graphic data were not stored by individual respondents within the sample closely resembles the U. bSome data sources combined these groups. 6a) and MOR (now HRV Step B. we changed its name from Less than 12 56 14 16 HEV to HRV. This examination led us to reduce the importance of AG African American 51 13 12 (now HRV Step B.. 1997. All of us provided which participants were lost when we eliminated the records with . Of these records. N = 389. & Braff. eth- census data. 4b in Table 2). Jansak. 1999. COP = cooperative movement. High HEV scores were in the negative. and education. but high school educated indi. & Gowri. 1992. 2001. As evident in Step C. Viglione. or undesir. aGiven in percentages. tions. 4a).1 for this sample. Asian American 18 5 4 European American 259 67 72 amining these responses also revealed that Anatomy Content Hispanic American 42 11 11 (An) should be addressed in the HRV algorithm (see HRV Native American 2 1 1 Step B. NA = not available. Offendersc 50 13 able direction. 1999. Exner et al. Ta. Numerous individuals We assembled a sample of 389 Rorschach records from eight administered these records at many different sites. Good Hu. Meyer. The data available in Table 6 suggest that the nicity. 1995. These eight groups were selected to represent a maximized external validity across examiners and institu- diverse range of impairment and administration contexts. Native American/otherb 3 1 NA Missing data 14 4 To distinguish the new algorithm from the original HEV Education (years) and to name it more accurately. we incorporated cul- tural and ethnic diversity by matching each of the eight 1 groups as closely as possible to the 1997 to 1998 estimated Some of the subgroup data included only summaries for age. to justify a simple difference score for the HRV calculation. DeLucas. Green. Accordingly. We also hoped to confirm Rorschach records from our laboratories and previous stud- that the GHR and PHR distributions are sufficiently similar ies (Cassella. the direction Community volunteers of scoring for the HRV was reversed relative to the HEV. population characteristics computerized Rorschach summary files but instead were reported in terms of age and ethnicity.5 (PHR). HEV = Human Experience Variable. cIncludes 6 offenders with severe psychiatric disorders. Outpatient 50 13 Inpatient 50 13 Patients with schizophrenia 51 13 HRV DEVELOPMENT STUDY Note. and a raw value difference score (rather than a > 16 19 5 8 weighted difference score as with the HEV). CS = Comprehensive System. 12 107 28 35 13 to 15 96 25 25 man Representation (GHR) and Poor Human Representation 16 45 12 16.

Table 7 contains the descriptive statistics for the HEV and ical impairment and approaches to the testing. Consistent with our goals. aMaximum = 10. As expected. 1. the distributions are ing that GHR may be constrained by the limited number of po- highly similar from the 0 through 95th percentiles.58 3. 1996). PHE = Poor Human Experience. GHR = Good Human Representation. the data and analyses in this article reveal that the There is one dissertation (McGlone.25 2.28 . The disparity in maximum GHR able. plates (Exner et al. PHR = Poor Human Representation.24 2.43 2.40 4.38 1 1 75th percentile 4 6 5 4 4 1. although the differ- scores are associated with healthy interpersonal perception ences were in the expected direction. These are somewhat confounded groups in that the au- ation.70 VIGLIONE. 0 0 0 0 0 –6. An examination of Table 7 reveals that ing that there are some very high values for PHR. In such a large sample with great diversity of psycholog. relatively equal proportions of individuals have PHR scores of 0. Ephebophile is a means and distribution are more nearly equal. If this unexpected inde- with schizophrenia. a group that presumably has an intact under- with healthy interpersonal perception and functioning.41 .70 –6 –5 25th percentile 2 1 2 1 1 –1. There were no significant HRV scale is reversed relative to the HEV.80 3. like the distribution for the former PHE vari. the GHR and PHR pedophile and ephebophile priests (N = 79). This cap was employed to explore the effects of the tain all data points but change all values greater than 10 to 10) skew with PHR relative to the maximum score of 10 found so as to maximize the similarity between GHR and PHR distri- with the GHR. MEYER.20 –19 –10 5th percentile 0 0 0 0 0 –3. re- of 10.45 Kurtosis .32 . N = 363. there seems to be a ceiling of about 10 for GHR suggest- to the GHE and PHE means.26 3.711 –1. Such a tactic would minimize the negative skew in the HEV and HRV and their subcomponents. and PHR scores produces a moderate degree of skew in the mum score of 25 in this sample.10 5 5 Max.88 . High HRV scores and low HEV scores are associated with more positive interpersonal perception. formation for the sample of 363. PHR is not capped HRV. whereas low HEV scores were associated with priests. HRV = Human Representational Variable. GHR was relatively high compared to A number of psychometric issues are worthy of consider. As might be expected and functioning. It has a tail at the high end and a maxi. HEV = Human Experience Variable. pendence holds up under further research. EXNER TABLE 7 Descriptive Data for the Old HEV and the New HRV HRV With Variable GHE PHE GHR PHR PHR–10a HEVb HRVb PHR–10a M 2. standing of others. Accordingly.80 8.07 .e. The great majority of the records Within the sample presented in Table 8. On the other the GHR and PHR means are more nearly equal as compared hand.95 3.09 2 2 95th percentile 6 9 7 8. In fact.61 7. a large study with Roman Catholic clergy as re- With the original HEV.22 Skew . distributions. The GHR distribution is comparably normal with relatively few (5. There are other small differences in the GHR and PHR at 10 in this table.72 9 9 Note.79 3.76 3. 2001). GHE = Good Human Experience. The last column in Table 7 composite HRV (Curran. & Finch. whereas the offender groups came from institutional settings and less than 14 responses and thus cannot present the demographic in- were tested as a part of formal clinical evaluations. thor tested the control group as research volunteers. which justifies term used to identify individuals who have molested adoles- using a raw score difference for computing the HRV (HRV = cents rather than prepubescent children.71 2.09 –. GHR and PHR . The PHR tential good human forms identifiable in the 10 Rorschach distribution. there were more PHE responses than spondents.38 . PHR. is highly skewed. DISCUSSION 3 in our sample (15% to 17%) so that most individuals give three or fewer PHR responses. GHR and PHR with fewer than 14 responses came from the sample of individuals are not correlated with one another. Table 8 provides correlations between the butions. presents the HRV with the PHR capped at a maximum score in research it might be advisable to truncate PHR at 10 (i. 2001) using the cur- HRV has improved psychometric qualities relative to the HEV.93 .89 3. 9 27 10 25 10 7.48 SD 1.2 %) zero values. rent HRV.96 3. This study contrasted control priests (N = 80) to GHE responses. Now high HRV HRV differences between these groups.25 3.07 9.43 Min. it is not surpris- HRV components. JANSAK. or.43 –.70 –1 –1 Mdn 3 3 3 3 3 –. bThe HRV is scored in the opposite direction of the HEV. West. In contrast..59 2. 2. It should be noted again that the direction of the study are presented in Table 3. The data from this GHR – PHR).44 1.. PERRY.19 .

S.42 . and suggestions. and FSIQ scores. This is consistent with the Note. and protocol complexity as potential moderator aminer. and describes people. Content. illogical. RORSCHACH HRV 71 TABLE 8 eralization is further supported by the fact that the HRV en- Correlations Between HRV and HEV compasses many response dimensions including Form and Subcomponents Quality. To put the strength of these corre. and Ann DeLucas . As with other Ror- sentations have been measured in many different ways in this schach variables. These high correlations support the position that the HRV should be in. social interactions. 1996. one would expect stronger relationships research. Steve Hickman. This gen. The first number listed is the Pearson correlation coefficient point of view (Viglione. and relationships. More recent research (e. and Special Scores. the respondent conceptual- HRVa .04/. 2000) and the qualitative anal- tional responses.–09 human experience.88/. 1999) that the most valid Rorschach and the second number is the Spearman rank order correlation coefficient. unrealistic. Across the older and newer versions of the WAIS. and PHE . Thus. N = 363. 1993. human responses and benign or cooperative in- and low HEV scores are associated with more positive interpersonal teractions. HEV = Human havior embodied in the response process. ligence Scale–Revised and the WAIS–III for VIQ. Also.77 the interpretation of the HRV. and relationships manifested in accurate.07 . Julie Bussman. and Thomas Boll for their expert observations The results in Tables 7 and 8 were generated with a psy. population so that the results should be and Suzanne O’Brien for their help in preparing this article applicable to diverse ethnic groups and settings. other. ous and under conditions in which the respondent has to lations in context. Accordingly. one elaborates the interpretation of the HRV interpersonal perception organized in a way that is not avail. Good Human Experience.87 to .92 –. for example.59 –. re. Debra Cha. and PHE scores likely would be strongest when situational cues are ambigu- (rs range from . High HRV scores ical. Burns.43/–. dam- aged.47/. log- aThe HRV is scored in the opposite direction of the HEV.80/. This is a robust finding in that human repre. it is not quality of human and interpersonal perceptions and represen- clear whether the interpersonal implications of a protocol tations. HRV. Viglione. terpreted like the HEV. sentations or perceptions.63 –.. Irving Wiener. lar to the U.96/. Populars. the alternate forms reliability coefficients range from ACKNOWLEDGMENTS .g. intact. PHR . HRV = Human Representational Variable. confused. visualizes. In turn. would be the tionships. Burns & with (a) behaviorally valid and perceptually based criteria Viglione. 1999. PIQ. GHR responses are Experience Variable. and to Elizabeth Green. ternal events.86 to . GHE. 1990. the total number of responses in the ysis of the Rorschach responses and relationship with the ex- protocol (R). Bridgett Scimone.85/–. heterogeneous dimensions.51/. the HRV summarizes the overall acteristics in different groups and situations. able from other CS Rorschach measures. Thanks to Phil Erdberg. Mark Hilsenroth. specific description of interpersonal perception and related- The HRV provides summary statistical information about ness. that is.94 –. Todd Pizitz. 1999). many of Variable GHR PHR HRV GHE PHE which are well represented in previous positive research findings. Ini- response qualities are associated with interpersonal function.13/.56 izes. the implicit understanding of people and rela- with GHR to PHR scores of 1 to 1. GHE = perceptions or representations of positive schema of self. this behavior is the foundation of HEV –. Viglione & Hilsenroth. Interpersonal Cluster (Exner. Asso- As expected.88/–.76/–.02/–. GHR = Good Human Representation. Joseph Obegi. aggressive.96).02/. one would use additional Rorschach data in terms of the CS search should explore the total number of human representa. and interpretations of people. PHR are negative or problematic perceptions or perception. it should be noted that they are extremely rely on himself or herself to make sense of external and in- similar to the correlations between the Wechsler Adult Intel. Extensive research Further elaboration of interpretations associated with in- indicates that a wide variety of Rorschach human content and dividual HRV score levels awaits additional research.63/–. based on additional data.87 –. to arrive at a more variables. HEV possessed similar interpretive qualities.00 To produce HRV responses. chiatrically diverse sample that was demographically simi. in addition to collateral data. representations as manifested in distorted. Along these broad and GHR and PHR may be related to different interpersonal char. ciations between the HRV and interpersonal criteria most related with the corresponding HEV. human intentions. PHR = Poor Human Representation. tial efforts should address the association of HRV with ing and psychological health (Stricker & Healy. Michael Cassella. realistic. same as a protocol with scores of 5 to 5. We also thank Nicole Taylor. also see Table 1) suggests that the original in comparison to (b) self-report criteria (Meyer & Archer. Viglione.94 (Psychological Corporation. 1997). 2001).77 GHE . 2001.71 –. representations. and PHR scores are highly cor. PHE = Poor Human Experience. or malevolent repre- may ultimately be interpreted or researched individually. interpretations are those that closely resemble the actual be- The corresponding HEV/HRV correlations are underlined. Within the interpretation of a Rorschach record.12 .66 . perceptions. GHR.76/.

(1967). (1976). The convergent validity of MMPI and Rorschach Brinderson. (1994). Ego Impairment Index: A Rorschach study with navy wives.. Journal of Personality Assessment. some of which Haller.. ment. (1993). K. A.. 31. Montemagni. Psycholog- in heterogeneous psychiatric patients. Exner. schach psychodiagnostic test. W. 59. Unpublished doctoral disserta- Blatt. EXNER who contributed data from their dissertations. 4. NC: Rorschach Workshops.. 1. lic priests: Characterization and analysis. A va- Adrian. Journal of Personality Assessment. S. Rorschach response. Assessing object representations Perry. 319–334. Califor- Curran... M. 64. The Ego Impairment Index control and affect modulation. C. C. & Hilton.. E. Hart. A validation study of Rorschach indicators of impulse Perry. Perry.). (1995). & Holmstrom. Object-representations and object-relationships in among opiate addicts. (1988). Journal of Personality Assessment. L. Fritsch. & Lerner. & Glick. Journal of Personality Assessment. M. Asheville. Journal of Personality Assessment. S. 55. The reliability of the Rorschach depressive indices in ma. M. (2001). Psychological Assessment. Apparent nonpatient psychopathology on the Ror.. Information-processing deficits and Exner. (1989). 72. 8. L..(4). A primer for Rorschach interpretation. E. B. California School of Professional Psychology. Perry. & Archer. Unpublished male late-life schizophrenia patients and its relationship to quality of life. 63. Hibbard. Unpublished doctoral dissertation. E. Schimek. M. West. B. Unpublished doctoral dissertation. 67.. Journal of Personality Assessment. L.. San Diego. S. Journal of Abnormal Psychology. and mediating factors. The quality of paternal attachment and violent behav. American Journal of Psychiatry. S. B. J. (1982). D. & Nash. W.. The Rorschach: A Comprehensive System: Vol. California School Of Professional Psychology. Hibbard. L. D. . 148–166. ality disorder. & Handler. Cassella. sic foundations (3rd ed. and schizophrenia: A validation study. 853–858. M. 62. J. Psychological Methods. P. J... Schaible. sessment. Amphetamine on Rorschach measures in normal sub- Journal of Personality Assessment. The Rorschach Comprehensive System Depression In- Blake. R. Perry. Hillman. Minassian.. 1–35. C. (1995). Biological Psychiatry. E. NC: Rorschach Workshops. & Braff. J. Journal of Personality Assessment. A. G. San Diego.. McDougall. A multimethod ap. J. 16–29. velopment and psychopathological impairment of the concept of the ob. A.. B. A five-year follow-up on Sciara. S. P. B. & Feldman. Journal of Projective Techniques & Personality As- Blatt. C. 13. (1999). & Braff. San Diego. Assessment. The Ego Impairment Index as a predictor of outcome in a lished doctoral dissertation. San Diego. 45–57. The Rorschach texture response: A conceptual vali. Alliant International Univer- nonpatients. L. Viglione. Jr. MEYER. 87–104. (1990). & Braff. The Rorschach Ego Impairment Index lidity study of two projective object representations measures. (1983).. D. Journal of Personality Assessment. Brenneis. styles on both methods and a reexamination of simple Rorschach re- California School of Professional Psychology.. San behavior. Green. (1996). J. Jenkins. 487–501. (1996). (1999). 486–502. 40. ual responses. G. (1995). J. C. (2000). Journal of Personality Assessment. Object relations theory and the Rorschach: An exami- jor depressive disorder. J. 64. Ingham. (1994). fessional Psychology. San Diego. Predicting adaptation to spousal separation with the Auslander. et al. The psychological assessment of object tion. & Viglione. J. depression heterogeneity. San Unpublished doctoral dissertation. tics to nonnormality and specification error in confirmatory factor analy. 17–24. (1994). 92–99. sponse frequency. J.. M. S. iors: A Rorschach study of convicted military male offenders. A. Differentiation of personality types Mayman.. CA. & Viglione. H. D. (1985). 399–413. Netter. Jr. D. jects. Netter. J. et al. D.. Fowler. 408–414. E. R. Journal of Personality System (5th ed. B. M. The relationship of infor. Blatt. Perry.. Jr. 363–367. D. Meyer. Stillwater. Self-delineation and dex. J. S.. J. C. Cadenhead. M. L. B. Sexual disturbances. D. Unpublished doctoral dissertation. 456–465. dissertation. The robustness of test statis. B. REFERENCES 116–132.. G. depressants. Sprock. melancholic depressed populations treated with tricyclic antidepressants. (1995). R. Humphrey. California School of Professional Psychology. JANSAK. (1991). 65.. J. California School of Pro. A. (1997). Hilsenroth. G. & Viglione. Hilsenroth.. Diego.72 VIGLIONE. nia School of Professional Psychology.. J. Asheville. D. N. Picker. (2000). & Kaser-Boyd. PERRY. San Diego.. California School of Professional Psychology. San Diego. Scales: An extension using profile scores to define response and character schach: Disturbance or disinhibition? Unpublished doctoral dissertation. San Diego. W. D. Journal of Personality Assessment. doctoral dissertation. McGlone. Jansak. (1993). Journal of Personality Assessment. C. (1992). Ba. 364–373. 516–521. J. Normal de. J. Burns. I. (1995). nia School of Professional Psychology. A Rorschach workbook for the Comprehensive the temporal stability of the Ego Impairment Index. Thought disturbance differences between male and fe. Morgan. McDougall. 49. 54. D. (2001). E. 151. and object representation in system. The Rorschach Human Experience suppressive testing contexts: A study of an extended aggression scoring Variable. (2003). 432–439. & Viglione. J. J. H. Colligan. W.. W. The reliability of Rorschach variables for inpatients presenting symptoms of depression and/or helplessness. thought disorder in schizophrenia.. 7. K. & Berman. E. Rorschach sex- mation-processing deficits and clinical symptoms in schizotypal person. Califor. (2001). Ritzler.). L. An object relations study: Relationship between Ror. chology. San Diego. (1992). L. ical Assessment. W. D. San Diego. (1984).. E. The hard science of Rorschach re- schach Human Experience variable and interpersonal relatedness among search: What do we know and where do we go? Psychological Assess- nonpatients. F. R. (1996). W. A. Meyer. J... & Finch. Unpublished doctoral International University. schizophrenia on the Rorschach. Diego. Haller. California School of Professional Psy- Exner. R. (1990). Sexually offending and non-offending Roman Catho- ject on the Rorschach. DeLucas. The Ego Impairment Index as a predictor as a continuous variable: A modification of the concept of the object on of outcome in melancholic depressed patients treated with tricyclic anti- the Rorschach scale.. M. K. J. Dimensions of personality organization as pre- dictors of teenage pregnancy risk. A. 112–118... W. 47. C... California School of Professional Psy. Psychological Assessment. N. Unpublished doctoral dissertation. 1. Oklahoma State University. Unpublished doctoral dissertation.. Unpub. N. (1990). The detection of faking schizophrenia in the Ror- dation study. S. D. ogy. S. and the role of self-schema. Journal of Personality Assessment.. 7–28. were used in this study. W. (1993). Aggressive Rorschach content in expressive and Burns. 530–536. 165–175. Perry. L. chology. (1995). 85. Unpublished doctoral dissertation. Unpublished marital interaction: The Rorschach predicts structural analysis of social doctoral dissertation. & Exner. proach to assessing dependency: The early memory dependency probe. Jr. New York: Wiley.. J. Jr. 56.. L. interpersonal relatedness.. California School of Professional Psychol.. E. G. (1999). United States nation into the impaired perception of humanness. Metts.. W. representation. sity.. (1996). San Diego. & Viglione. An empirical study of malingering sis. 52.

& Gowri. CA. Alliant International University Viglione. & Ortiz.. M. B. The Rorschach: Facts. J. K.. G.. April). Viglione ity Assessment. 11. Viglione (Chair). and ment. T.. Journal of Personality Assessment. 2003 . California School of Professional vancing scientific discourse in the controversy surrounding the Compre- Psychology. Validity of the Rorschach Mutuality of Auton- omy Scale: A replication using excerpted responses. & Grolnick.. Weiner (1996). S. D.. adult. San Diego. T. S. 41. Two questions San Antonio. & Hilsenroth. 115–129. (1997. Gaudiana. 10455 Pomerado Road September).. Barker. M. Viglione. (1977). of Consulting and Clinical Psychology. Strauss. R. New York: Wiley. Brager. (1966). M. Nezworski. Differential diagnosis of battered women through psy. 13. W.. 49. J. and priest sexual offenders. M.. W. Urist. A review of the empirical literature. Journal of Personal- Donald J. M. J. 6–12.. & Healy. (1982). Revised February 3. G. In D. RORSCHACH HRV 73 Psychological Corporation. The Rorschach Test and the assessment of object relations. (1999). Descriptive and predictive implications of Rorschach and San Diego.2001 Conference on Family Violence.edu treatment. Viglione. I. (1996). & Shill. CA. Clinical and risk-management impli- cations of psychological assessment research with adolescent. J. 450–454. B. (2001). 251–265. hensive System for the Rorschach. Wood.. Ad- der? Unpublished doctoral dissertation. 3–9. San Diego. (2001. 6. W. J. at the Society for Personality Assessment. (2001). Journal of Personality Assess. Stejskal.. A review of recent research addressing the utility of 201 Daley Hall the Rorschach. & Garven. R. Psychological Assessment. chological testing: Personality disorder or post-traumatic stress disor. 2... (1985). CA 92131–1799 other psychological assessment data from adolescent sexual offenders in E-mail: dviglione@alliant. Viglione. Weiner. Psychodiagnosis in schizophrenia. B. future. San Diego. Moore.. about Rorschach norms: Cultural issues and divergence.. Projective assessment of object relations: 369–378. Methodological issues in evaluating Rorschach validity: A com- Urist. J. fiction. J. J. Avery. Paper presented Ryan. WAIS–III & WMS–III technical manual. & West. 46. Psychological Assessment: A Journal Wood. Garven. 76. T. (1997). 219–230. (1999).. D. Nezworski. Stejskal. R. (1990). J. M. Bass.. Journal of Personality Assessment. Stricker. and Ganellen (1996). Psychological Assessment. D. D.. J. M. J. A. J. Assessment. TX: Author. 452–471. A Rorschach assess. S. ment of children’s mutuality of autonomy. M. R. Symposium conducted at the 6th International Received December 2. S. J. S. ment on Burns and Viglione (1996).