This article was downloaded by: [Deakin University Library] On: 13 March 2015, At: 14:45 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Journal of Personality Assessment Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/hjpa20
Estimation of MCMI DSM-III Axis II Constructs From MMPI Scales and Subscales Karen L. Zarella , James M. Schuerger & George H. Ritz Published online: 22 Jun 2011.
To cite this article: Karen L. Zarella , James M. Schuerger & George H. Ritz (1990) Estimation of MCMI DSM-III Axis II Constructs From MMPI Scales and Subscales, Journal of Personality Assessment, 55:1-2, 195-201, DOI: 10.1080/00223891.1990.9674058 To link to this article: http://dx.doi.org/10.1080/00223891.1990.9674058
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
JOURNAL OF PERSONALITY ASSESSMENT, 1990, 55(1&2), 195-201 Copyright o 1990, Lawrence Erlbaum Associates, Inc.
Estimation of MCMI DSM-111 Axis I1 Constructs From MMPI Scales and Subscales Karen L. Zarrella and James M. Schuerger Downloaded by [Deakin University Library] at 14:45 13 March 2015
Cleveland State University
George H. Ritz Cleveland, OH
Wiggins, Harris and Lingoes, and Serkownek Minnesota Multiphasic Personality Inventory (MMPI) scores were used to predict Millon Clinical Multiaxial Inventory (MGMI) scores in a 100-patient sample. Equations from the first sample were cross-validated on a sample of 212 inmate subjects. We conclude that scores on 19 of the 20 MCMI scales can be successfully predicted by the Wiggins, Harris and Lingoes, and Serkownek subscales of the MMPI. In further cross-validation,the equations were used to predict the Morey, Waugh, and Blashfield MMPI composites for the prison sample, again with strongly positive results. The results appear quite promising for the estimation of personality disorder constructs from MMPI scales and subscales.
There are two contexts for this research. The first is the scattered research o n concordance among various psychometric instruments. T h e second is the recent creation of psychometric scales aligned with Personality Disorders as defined in the Diagnostic and Statistical Manual of Mental Disorders (3rd ed. [DSM-UI], American Psychiatric Association, 1980). Past research i n the first of these areas has included re dictions between the MMPI and the Sixteen Personality Factor Questionnaire (16PF; Cattell & Boulton, 1969; Cattell, Eber, & Tatsuoka, 1970), and between t h e MMPI and California Psychological Inventory (CPI; Gough, 1974). These studies permit the estimation of scale scores o n one inventory from those o n another by providing a set of cross-inventory prediction equations via multiple regression procedures. In the second area we may list t h e Millon Clinical Multiaxial Inventory (MCMI) and MCMI-I1 (1983, 1987) and the Morey, Waugh, and Blashfield (MWB; 1985) scales for the MMPI. T h e
196
ZARRELLA, SCHUERGER, RITZ
main purpose of this study is to further these efforts by establishing that MMPI scales and subscales can be used to estimate, as markers for DSM-I11 Axis 11 constructs, the relevant scale scores of the MCMI, and that these same estimates correlate substantially with the MWB scales.
Downloaded by [Deakin University Library] at 14:45 13 March 2015
VARIABLES The MCMI (Millon, 1983), although a comparatively new psychometric tool, has been notable in its applicability to the assessment of personality disorders, particularly to the DSM-I11 Axis 11formulations. The MCMI was constructed using a relatively new procedure involving a sequential approach involving three stages: theoretical-intuitive generation of items, internal-structure revisions through internal-consistency measures, external-validation procedures involving a solid distinction between diagnostic groups and reference groups, and additional cross-validationprocedures. Twenty separate scales are provided under three general headings encompassing relatively enduring personality characteristics and severe clinical states: Basic Personality Styles (Schizoid, Avoidant, Dependent, Histrionic, Narcissistic, Antisocial, Compulsive, and Passive-Aggressive), Pathological Personality Syndromes (Schizotypal, Borderline, and Paranoid), and Symptom Disorders (Anxiety, Somatoform, Hypomanic, Dysthymia, Alcohol Abuse, Drug Abuse, Psychotic Thinking, Psychotic Depression, and Psychotic Delusions). The variables of particular interest in this study, aligned with DSM-111 Axis I1 disorders, are presented in Table 1. Also presented in Table 1 for each disorder is the 3 x 3 table of intercorrelations among the three estimates of each disorder construct-one from the MCMI, one from the MWB scales from MMPI items, and one from the MMPI scale composite generated in this study. The MMPI, although originally created by a contrasted-groups strategy to identify mainly Axis I disorders, has subsequently yielded a great many content scales (Graham, 1977). Among these t~ The MWB are the relatively new MWB scales for DSM-Ill ~ e r s o n a l idisorders. personality disorder scales, derived through a combined rational/empirical strategy, encompass all 11 personality disorders and appear to represent an alternate conceptualization of the DSM-III personality disorders.
PARTICIPANTS Two separate samples were used in this study, in many ways quite different from each other, but similar in that both are clinical samples appropriate for these instruments. One hundred patient files were randomly chosen from a psychologist's private practice for the primary sample, which served as the basis for
PREDICTION
197
TABLE 1 3 x 3 Correlation Table of MCMI, MWB, and MMPI Composite Variables Patient Sample
DSM-I11 Axis I1 Schizoid Avoidant
Downloaded by [Deakin University Library] at 14:45 13 March 2015
Dependent Histrionic Narcissistic Antisocial Obsessive-Compulsive Passive-Aggressive Schizotypal Borderline Paranoid
Variables
MCMI
MWE
Prisoner Sample
MCMI
-
MWB
MCMI MCI MWB SZD MMPI Compl MCMI MC2 MWB AVD MMPI Comp2 MCMI MC3 MWB DEP MMPI Comp3 MCMI MC4 MWB HST MMPI Comp4 MCMI MC5 MWB NAR MMPI Comp5 MCMI MC6 MWB ANT MMPI Comp6 MCMI MC7 MWB CPS MMPI Cornp7 MCMI MC8 MWB PAG MMPI Comp8 MCMI S MWB STY MMPI Comp9 MCMI C MWB BDL MMPI ComplO MCMI P MWB PAR MMPI Compll
regression equations. Of the sample chosen, 51% were men and 49% were women. The mean age of the subjects was 36.7. For cross-validation purposes, 212 inmates were tested at an Eastern correctional institute for women. The mean age of the subjects was 29.7. Of the prison sample, 66% of the subjects were Afro-American, 33% were Caucasian, and the remaining 1% consisted of subjects with either a Puerto Rican or Indian heritage. The average educational level of the subjects was 11.3 years.
Downloaded by [Deakin University Library] at 14:45 13 March 2015
PROCEDURE For the private-practice sample, patients were routinely administered a standard battery of psychological assessment inventories at various stages in the process of psychological counseling. All patient file information was numerically coded to protect the identities of the patients. For the prison sample, inmates incarcerated at the institution between July 1985 through December 1987, who either requested psychological counseling or were referred through another professional at the institution, were administered the psychological assessment inventories upon entry. The psychological instruments were administered by a licensed correctional psychologist who provided each subject with a coded identification number in order to maintain the anonymity of each inmate. Inmates received a verbal briefing thoroughly explaining the purpose of the research.
RESULTS A two-stage stepwise multiple regression analysis was run under a set of specific conditions. The MCMI scores served as the dependent variables. The first stage involved entering only the Wiggins content scales (SOC through HEA); the second stage entered the Harris and Lingoes subscales (Dl through MA4), as well as the Serkownek subscales (MF1 through 96). The main clinical scales of the MMPI were not utilized because of their heterogeneity of content and their greater emphasis on Axis I clinical-syndrome disorders. Because they largely circumscribe the main content areas of the MMPI, the Wiggins scores were selected for first inclusion. Subsequently, the Harris and Lingoes and the Serkownek subscales were allowed to enter, to pick up lesser content areas that might contribute to MCMI variance. An additional reason for choosing the Wiggins, Harris and Lingoes, and Serkownek variables was that, because of their inclusion in the commercial scoring routines, they have some likelihood of being familiar to users. ) for in the target MCMI scales The percentage of variance ( R ~accounted ranged from .44 to .78, with 16 of the 20 scales accounting for greater than 50% of the variance. (Those interested can obtain a comprehension listing of regression analyses by contacting James M. Schuerger.) The cross-validation phase involved a correlation of the composites with the actual MCMI scores for the prison sample. The analysis revealed substantial ~redictabilityfor 17 of the 20 MCMI scales. Scales ranged from .50 to -77, all at a significance level of p < .001. Scales H (Somatoform) and P (Paranoid) exhibited somewhat lower coefficientswith correlations of .38, p < .001 and .29, p < .001, respectively. Scale PP (Psychotic Delusion) fared rather poorly with a coe6cient of .18, P < .01.
As a supplementary analysis, the composites from the MCMI Axis 11 equations for the patient sample were again used in the prison sample. However, this time the MWB alternate estimates of the Axis I1 disorders were used as criteria. All 11 of these concept-validation coefficients were significant (P < .001), and 8 of them were substantial (> .50).
Downloaded by [Deakin University Library] at 14:45 13 March 2015
DISCUSSION The first aspect of this investigation addressed the possible relationship between the MMPI subscales and MCMI scale scores. This study has clearly shown that composites from one psychometric tool (specifically the lvIMPI subscales) correlate with scales from another instrument (here, the MCh4I) at about the level of short-term (2 months to 1 year) test-retest reliability (Schuerger, Tait, & Tavernelli, 1982; Schuerger, Zarrella, & Hotz, 1989). A second aspect, perhaps of more importance, is the potential predictive contribution of the MMPI subscales in relation to the 11 DSM-I11 Axis II personality disorders, as exemplified through the utilization of the MCMI or MWE; scales. That is to say, can the MMM subscales (Wiggins, Harris and Lingoes, and Serkownek) play a functional role in the identification of the 11 Axis ][Ipersonality disorders? As a first step toward understanding this issue, a theorietical examination of the constructs afforded through the 11 regression equations provides some interesting insights. For ease of perusal, Table 2 is provided to illustrate the specific MMPI subscales contributing to the construct makeup of only the Millon Axis 11scales. As can be seen, the patterns of M I P I variables related to Axis 11 constructs is theoretically consistent with the empirical makeup of the Axis 11disorders. For example, in the second analysis, the dependent variable MC2 (Avoidant) was significantly predicted by Wiggins scales SOC (Social Maladjustment), DEP (Depression), and MOR (Poor Morale). In addition, Harris and Lingoes subscales provided the predictors SC2 (Lack of Ego Mastery-Cognitive and Conative), SC2C (Defective Inhibition), and PD4 (Social/Self-Alienation). When merged, these variables are consisltent with behavioral descriptions of the avoidant personality. The third analysis, concerning the cross-validation of MWB and MMPI composite variables, provided further substantiation of our results. There is one apparent anomaly in our data which bears further scrutiny. In particular, scale MC7 of the MCMI and scale CPS of the MWB were found to correlate strongly in the negative direction in both the patient and prisoner samples. This is in spite of the fact that each scale was drawn from the DSM-I11 Axis I1 classification of the obsessive-compulsive personality. It appears that these two scales may actually be measuring quite diverse aspects of this Axis I1 disorder. Further investigation of the conceptual basis for these two scales would be necessary to resolve the apparent conflict.
Downloaded by [Deakin University Library] at 14:45 13 March 2015
TABLE 2 MMPI Subscales Relevant t o Prediction of AXIS I1 MCMI Scales Disorder Composite Compl-Schizoid Comp2-Avoidant Comp3-Dependent Comp4-Histrionic Comp5-Narcissistic Comp6-Antisocial Comp7-Compulsive Comp8-Passive-Aggressive
Comp9-Schizotypal Comp 10-Borderline Compl 1-Paranoid
Wiggins Content Scales SOC,DEP SOC,DEP(-),MOR MOR,HOS(-)
SOC(-),HOS,FEM,REL(-),HEA(-) SOC(-),DEP(-),REL,HOS MOR(- ),HOS,HYP SOC,MOR(-),HOS(-) DEP,HYP SOC,DEP,MOR,HOS(- ) SOC(-),MOR SOC(-),REL,HYP,AUT
Note. Contribution to multiple prediction is positive except where indicated (-).
Harris Subscales PAl,PA2(-),SC2A,MA3 PD4,SC2,SC2C(-) Dl(-),SC2A HY2 PD1(-),PD3 PA3(-) PD2(-),PD4B(-),SC2C PD4B PD4,MA4(-) MAZ(-)
Serkownek Subscales -
S14 MF2 MF2 MF4,SI1(-) MF2(-) MF2
-
MF5 SI6(-)
PREDICTION
201
Downloaded by [Deakin University Library] at 14:45 13 March 2015
Our research has illustrated the usefulness of the cross-correlation approach for examining the congruence of independently developed psychometric tools. Such investigations can provide a consistent set of prediction equations which could aid clinicians in relating results from two or more inventories. Of equal importance is the ability to assess the consistency of apparently equivalent personality disorder construct measures. Differences observed in crossvalidation studies can provide a starting point for examining subtle but substantial differences between seemingly related construct measures. A more thorough understanding of such differences is necessary if combined batteries are to be used to their full predictive capacity.
REFERENCES American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. Cattell, R. B., & Boulton, L. S. (1969). What pathological dimensions lie beyond the normal dimensions of the 16PF?A comparison of MMPI and 16PF factors domains. Journal of Consalting and Clinical Psychology, 33, 18-29. Cattell, R. B., Eber, H. W., & Tatsuoka, M. M. (1970). Handbook for the Sixteen Pemonality Fuctor Questionnaire (16PF) in clinical, educational, industrial and research psychology. Champaign, IL: Institute for Personality and Ability Testing. Gough, H. G. (1974). Estimation of locus-of-control scores from the California Psychological Inventory. Psychological Reports, 35, 343-348. Graham, J. R. (1977). The MMPI: A praaical guide. New York: Oxford University Press. Millon, T. (1983). Millon Clinical Multiaxial Inventory manuul (3rd ed.). Minneapolis: National Computer Systems. Millon, T. (1987). Manual for the MCMI-Il(2nd ed.). Minneapolis: National Computer Systems. Morey, L. C., Waugh, M. H., & Blashfield, R. K. (1985). MMPI scales for DSM-I11 personality disorders: Their derivation and correlates. Journal of Personality Assessment, 49, 245-251. Schuerger, J. M., Tait, E., & Tavernelli, M. (1982). Temporal stability of personality by questionnaire. Journal of Personality and Social Psychology, 43, 176-182. Schuerger,J. M., Zarrella, K. L., &Hotz, A. S. (1989). Factors which influence the temporal stability of personality by questionnaire. Journal of Persodity and Social Psychology, 56, 777-783.
James M. Schuerger Department of Psychology Cleveland State University Cleveland, OH 44 115 Received October 6, 1988 Revised June 9, 1989
Journal of Personality Assessment Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/hjpa20
Estimation of MCMI DSM-III Axis II Constructs From MMPI Scales and Subscales Karen L. Zarella , James M. Schuerger & George H. Ritz Published online: 22 Jun 2011.
To cite this article: Karen L. Zarella , James M. Schuerger & George H. Ritz (1990) Estimation of MCMI DSM-III Axis II Constructs From MMPI Scales and Subscales, Journal of Personality Assessment, 55:1-2, 195-201, DOI: 10.1080/00223891.1990.9674058 To link to this article: http://dx.doi.org/10.1080/00223891.1990.9674058
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
JOURNAL OF PERSONALITY ASSESSMENT, 1990, 55(1&2), 195-201 Copyright o 1990, Lawrence Erlbaum Associates, Inc.
Estimation of MCMI DSM-111 Axis I1 Constructs From MMPI Scales and Subscales Karen L. Zarrella and James M. Schuerger Downloaded by [Deakin University Library] at 14:45 13 March 2015
Cleveland State University
George H. Ritz Cleveland, OH
Wiggins, Harris and Lingoes, and Serkownek Minnesota Multiphasic Personality Inventory (MMPI) scores were used to predict Millon Clinical Multiaxial Inventory (MGMI) scores in a 100-patient sample. Equations from the first sample were cross-validated on a sample of 212 inmate subjects. We conclude that scores on 19 of the 20 MCMI scales can be successfully predicted by the Wiggins, Harris and Lingoes, and Serkownek subscales of the MMPI. In further cross-validation,the equations were used to predict the Morey, Waugh, and Blashfield MMPI composites for the prison sample, again with strongly positive results. The results appear quite promising for the estimation of personality disorder constructs from MMPI scales and subscales.
There are two contexts for this research. The first is the scattered research o n concordance among various psychometric instruments. T h e second is the recent creation of psychometric scales aligned with Personality Disorders as defined in the Diagnostic and Statistical Manual of Mental Disorders (3rd ed. [DSM-UI], American Psychiatric Association, 1980). Past research i n the first of these areas has included re dictions between the MMPI and the Sixteen Personality Factor Questionnaire (16PF; Cattell & Boulton, 1969; Cattell, Eber, & Tatsuoka, 1970), and between t h e MMPI and California Psychological Inventory (CPI; Gough, 1974). These studies permit the estimation of scale scores o n one inventory from those o n another by providing a set of cross-inventory prediction equations via multiple regression procedures. In the second area we may list t h e Millon Clinical Multiaxial Inventory (MCMI) and MCMI-I1 (1983, 1987) and the Morey, Waugh, and Blashfield (MWB; 1985) scales for the MMPI. T h e
196
ZARRELLA, SCHUERGER, RITZ
main purpose of this study is to further these efforts by establishing that MMPI scales and subscales can be used to estimate, as markers for DSM-I11 Axis 11 constructs, the relevant scale scores of the MCMI, and that these same estimates correlate substantially with the MWB scales.
Downloaded by [Deakin University Library] at 14:45 13 March 2015
VARIABLES The MCMI (Millon, 1983), although a comparatively new psychometric tool, has been notable in its applicability to the assessment of personality disorders, particularly to the DSM-I11 Axis 11formulations. The MCMI was constructed using a relatively new procedure involving a sequential approach involving three stages: theoretical-intuitive generation of items, internal-structure revisions through internal-consistency measures, external-validation procedures involving a solid distinction between diagnostic groups and reference groups, and additional cross-validationprocedures. Twenty separate scales are provided under three general headings encompassing relatively enduring personality characteristics and severe clinical states: Basic Personality Styles (Schizoid, Avoidant, Dependent, Histrionic, Narcissistic, Antisocial, Compulsive, and Passive-Aggressive), Pathological Personality Syndromes (Schizotypal, Borderline, and Paranoid), and Symptom Disorders (Anxiety, Somatoform, Hypomanic, Dysthymia, Alcohol Abuse, Drug Abuse, Psychotic Thinking, Psychotic Depression, and Psychotic Delusions). The variables of particular interest in this study, aligned with DSM-111 Axis I1 disorders, are presented in Table 1. Also presented in Table 1 for each disorder is the 3 x 3 table of intercorrelations among the three estimates of each disorder construct-one from the MCMI, one from the MWB scales from MMPI items, and one from the MMPI scale composite generated in this study. The MMPI, although originally created by a contrasted-groups strategy to identify mainly Axis I disorders, has subsequently yielded a great many content scales (Graham, 1977). Among these t~ The MWB are the relatively new MWB scales for DSM-Ill ~ e r s o n a l idisorders. personality disorder scales, derived through a combined rational/empirical strategy, encompass all 11 personality disorders and appear to represent an alternate conceptualization of the DSM-III personality disorders.
PARTICIPANTS Two separate samples were used in this study, in many ways quite different from each other, but similar in that both are clinical samples appropriate for these instruments. One hundred patient files were randomly chosen from a psychologist's private practice for the primary sample, which served as the basis for
PREDICTION
197
TABLE 1 3 x 3 Correlation Table of MCMI, MWB, and MMPI Composite Variables Patient Sample
DSM-I11 Axis I1 Schizoid Avoidant
Downloaded by [Deakin University Library] at 14:45 13 March 2015
Dependent Histrionic Narcissistic Antisocial Obsessive-Compulsive Passive-Aggressive Schizotypal Borderline Paranoid
Variables
MCMI
MWE
Prisoner Sample
MCMI
-
MWB
MCMI MCI MWB SZD MMPI Compl MCMI MC2 MWB AVD MMPI Comp2 MCMI MC3 MWB DEP MMPI Comp3 MCMI MC4 MWB HST MMPI Comp4 MCMI MC5 MWB NAR MMPI Comp5 MCMI MC6 MWB ANT MMPI Comp6 MCMI MC7 MWB CPS MMPI Cornp7 MCMI MC8 MWB PAG MMPI Comp8 MCMI S MWB STY MMPI Comp9 MCMI C MWB BDL MMPI ComplO MCMI P MWB PAR MMPI Compll
regression equations. Of the sample chosen, 51% were men and 49% were women. The mean age of the subjects was 36.7. For cross-validation purposes, 212 inmates were tested at an Eastern correctional institute for women. The mean age of the subjects was 29.7. Of the prison sample, 66% of the subjects were Afro-American, 33% were Caucasian, and the remaining 1% consisted of subjects with either a Puerto Rican or Indian heritage. The average educational level of the subjects was 11.3 years.
Downloaded by [Deakin University Library] at 14:45 13 March 2015
PROCEDURE For the private-practice sample, patients were routinely administered a standard battery of psychological assessment inventories at various stages in the process of psychological counseling. All patient file information was numerically coded to protect the identities of the patients. For the prison sample, inmates incarcerated at the institution between July 1985 through December 1987, who either requested psychological counseling or were referred through another professional at the institution, were administered the psychological assessment inventories upon entry. The psychological instruments were administered by a licensed correctional psychologist who provided each subject with a coded identification number in order to maintain the anonymity of each inmate. Inmates received a verbal briefing thoroughly explaining the purpose of the research.
RESULTS A two-stage stepwise multiple regression analysis was run under a set of specific conditions. The MCMI scores served as the dependent variables. The first stage involved entering only the Wiggins content scales (SOC through HEA); the second stage entered the Harris and Lingoes subscales (Dl through MA4), as well as the Serkownek subscales (MF1 through 96). The main clinical scales of the MMPI were not utilized because of their heterogeneity of content and their greater emphasis on Axis I clinical-syndrome disorders. Because they largely circumscribe the main content areas of the MMPI, the Wiggins scores were selected for first inclusion. Subsequently, the Harris and Lingoes and the Serkownek subscales were allowed to enter, to pick up lesser content areas that might contribute to MCMI variance. An additional reason for choosing the Wiggins, Harris and Lingoes, and Serkownek variables was that, because of their inclusion in the commercial scoring routines, they have some likelihood of being familiar to users. ) for in the target MCMI scales The percentage of variance ( R ~accounted ranged from .44 to .78, with 16 of the 20 scales accounting for greater than 50% of the variance. (Those interested can obtain a comprehension listing of regression analyses by contacting James M. Schuerger.) The cross-validation phase involved a correlation of the composites with the actual MCMI scores for the prison sample. The analysis revealed substantial ~redictabilityfor 17 of the 20 MCMI scales. Scales ranged from .50 to -77, all at a significance level of p < .001. Scales H (Somatoform) and P (Paranoid) exhibited somewhat lower coefficientswith correlations of .38, p < .001 and .29, p < .001, respectively. Scale PP (Psychotic Delusion) fared rather poorly with a coe6cient of .18, P < .01.
As a supplementary analysis, the composites from the MCMI Axis 11 equations for the patient sample were again used in the prison sample. However, this time the MWB alternate estimates of the Axis I1 disorders were used as criteria. All 11 of these concept-validation coefficients were significant (P < .001), and 8 of them were substantial (> .50).
Downloaded by [Deakin University Library] at 14:45 13 March 2015
DISCUSSION The first aspect of this investigation addressed the possible relationship between the MMPI subscales and MCMI scale scores. This study has clearly shown that composites from one psychometric tool (specifically the lvIMPI subscales) correlate with scales from another instrument (here, the MCh4I) at about the level of short-term (2 months to 1 year) test-retest reliability (Schuerger, Tait, & Tavernelli, 1982; Schuerger, Zarrella, & Hotz, 1989). A second aspect, perhaps of more importance, is the potential predictive contribution of the MMPI subscales in relation to the 11 DSM-I11 Axis II personality disorders, as exemplified through the utilization of the MCMI or MWE; scales. That is to say, can the MMM subscales (Wiggins, Harris and Lingoes, and Serkownek) play a functional role in the identification of the 11 Axis ][Ipersonality disorders? As a first step toward understanding this issue, a theorietical examination of the constructs afforded through the 11 regression equations provides some interesting insights. For ease of perusal, Table 2 is provided to illustrate the specific MMPI subscales contributing to the construct makeup of only the Millon Axis 11scales. As can be seen, the patterns of M I P I variables related to Axis 11 constructs is theoretically consistent with the empirical makeup of the Axis 11disorders. For example, in the second analysis, the dependent variable MC2 (Avoidant) was significantly predicted by Wiggins scales SOC (Social Maladjustment), DEP (Depression), and MOR (Poor Morale). In addition, Harris and Lingoes subscales provided the predictors SC2 (Lack of Ego Mastery-Cognitive and Conative), SC2C (Defective Inhibition), and PD4 (Social/Self-Alienation). When merged, these variables are consisltent with behavioral descriptions of the avoidant personality. The third analysis, concerning the cross-validation of MWB and MMPI composite variables, provided further substantiation of our results. There is one apparent anomaly in our data which bears further scrutiny. In particular, scale MC7 of the MCMI and scale CPS of the MWB were found to correlate strongly in the negative direction in both the patient and prisoner samples. This is in spite of the fact that each scale was drawn from the DSM-I11 Axis I1 classification of the obsessive-compulsive personality. It appears that these two scales may actually be measuring quite diverse aspects of this Axis I1 disorder. Further investigation of the conceptual basis for these two scales would be necessary to resolve the apparent conflict.
Downloaded by [Deakin University Library] at 14:45 13 March 2015
TABLE 2 MMPI Subscales Relevant t o Prediction of AXIS I1 MCMI Scales Disorder Composite Compl-Schizoid Comp2-Avoidant Comp3-Dependent Comp4-Histrionic Comp5-Narcissistic Comp6-Antisocial Comp7-Compulsive Comp8-Passive-Aggressive
Comp9-Schizotypal Comp 10-Borderline Compl 1-Paranoid
Wiggins Content Scales SOC,DEP SOC,DEP(-),MOR MOR,HOS(-)
SOC(-),HOS,FEM,REL(-),HEA(-) SOC(-),DEP(-),REL,HOS MOR(- ),HOS,HYP SOC,MOR(-),HOS(-) DEP,HYP SOC,DEP,MOR,HOS(- ) SOC(-),MOR SOC(-),REL,HYP,AUT
Note. Contribution to multiple prediction is positive except where indicated (-).
Harris Subscales PAl,PA2(-),SC2A,MA3 PD4,SC2,SC2C(-) Dl(-),SC2A HY2 PD1(-),PD3 PA3(-) PD2(-),PD4B(-),SC2C PD4B PD4,MA4(-) MAZ(-)
Serkownek Subscales -
S14 MF2 MF2 MF4,SI1(-) MF2(-) MF2
-
MF5 SI6(-)
PREDICTION
201
Downloaded by [Deakin University Library] at 14:45 13 March 2015
Our research has illustrated the usefulness of the cross-correlation approach for examining the congruence of independently developed psychometric tools. Such investigations can provide a consistent set of prediction equations which could aid clinicians in relating results from two or more inventories. Of equal importance is the ability to assess the consistency of apparently equivalent personality disorder construct measures. Differences observed in crossvalidation studies can provide a starting point for examining subtle but substantial differences between seemingly related construct measures. A more thorough understanding of such differences is necessary if combined batteries are to be used to their full predictive capacity.
REFERENCES American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. Cattell, R. B., & Boulton, L. S. (1969). What pathological dimensions lie beyond the normal dimensions of the 16PF?A comparison of MMPI and 16PF factors domains. Journal of Consalting and Clinical Psychology, 33, 18-29. Cattell, R. B., Eber, H. W., & Tatsuoka, M. M. (1970). Handbook for the Sixteen Pemonality Fuctor Questionnaire (16PF) in clinical, educational, industrial and research psychology. Champaign, IL: Institute for Personality and Ability Testing. Gough, H. G. (1974). Estimation of locus-of-control scores from the California Psychological Inventory. Psychological Reports, 35, 343-348. Graham, J. R. (1977). The MMPI: A praaical guide. New York: Oxford University Press. Millon, T. (1983). Millon Clinical Multiaxial Inventory manuul (3rd ed.). Minneapolis: National Computer Systems. Millon, T. (1987). Manual for the MCMI-Il(2nd ed.). Minneapolis: National Computer Systems. Morey, L. C., Waugh, M. H., & Blashfield, R. K. (1985). MMPI scales for DSM-I11 personality disorders: Their derivation and correlates. Journal of Personality Assessment, 49, 245-251. Schuerger, J. M., Tait, E., & Tavernelli, M. (1982). Temporal stability of personality by questionnaire. Journal of Personality and Social Psychology, 43, 176-182. Schuerger,J. M., Zarrella, K. L., &Hotz, A. S. (1989). Factors which influence the temporal stability of personality by questionnaire. Journal of Persodity and Social Psychology, 56, 777-783.
James M. Schuerger Department of Psychology Cleveland State University Cleveland, OH 44 115 Received October 6, 1988 Revised June 9, 1989
