Cognitive Imbalance and AntisocialPersonality Characteristics

Ä

Mark Snow and Steven ThurberBoise State University

A cognitive imbalance, in which intellectual functioning is elevated in theperformance area in comparison to verbal IQ, has been posited as anantecedent condition in relation to antisocial behaviors. The current inves-tigation was based on the notion of a developmental arrest in which ver-bal, analytical, controlling brain processes (analogous to verbal IQ) fail todevelop commensurately with the more impulsive actions mediated bythe motor areas of the cerebral cortex (analogous to performance IQ). Thesimple verbal IQ performance IQ discrepancy index used in prior studieswas reformulated as a causal theoretical model consisting of shared andunique performance IQ variance. The participants were 325 adults includ-ing 141 prison inmates. They were administered the Wechsler Adult Intel-ligence Scale-Revised (WAIS-R) and the Psychopathic Deviate (Pd) andMania (Ma) scales of the Minnesota Multiphasic Personality Inventory-2(MMPI-2). These were the manifest (measured) variables in the modeltested by means of structural equation modeling procedures. Several sta-tistical indices suggested an excellent model-data congruence. © 1997John Wiley & Sons, Inc. J Clin Psychol 53: 351–354, 1997.

The notion that crime and delinquency may be related to an imbalance in performance andverbal intelligence (e.g., performance greater than verbal) has been corroborated in severalstudies with juvenile offenders (Cornell & Wilson, 1992; Miller, 1987; Walsh, 1991). Onetheoretical explanation for this finding is that, relative to verbal IQ, elevated performancescores may represent a type of developmental arrest in regard to brain maturation. That is tosay, the performance IQ involves, in part, the functioning of the motor areas of the right cere-bral cortex. In normal human development the motor areas show early maturation, thus pro-viding adaptive requirements for a rapid responding to stimuli on the part of infants and toddlers.The usual developmental course of brain maturation may include a general adaptive slowing of

Correspondence concerning this article can be sent to Mark E. Snow, Department of Psychology, Boise State Univer-sity, 1910 University Drive, Boise ID 83725.

JOURNAL OF CLINICAL PSYCHOLOGY, Vol. 53(4), 351–354 (1997)© 1997 John Wiley & Sons, Inc. CCC 0021-9762/97/040351-04

responses to stimuli with the increasing involvement of a left hemisphere interpretive processes(Joseph, 1982). The arrested development hypothesis refers specifically to a failure of verbal,analytical brain processes to develop commensurately with the more impulsive, irrational behav-iors mediated by the motor areas. Put another way, the individual oriented toward delinquencyand criminal acts may, in some ways, have comparatively less mature brain functioning thanpersons not so oriented. This may be reflected in a measured discrepancy between performanceand verbal intelligence.

The typical approach in conducting research in this domain is simply to evaluate the rela-tionship between performance minus verbal IQ test scores and some measures of criminality(e.g., Petee & Walsh, 1987). However, the formulations of Joseph (1982) involve the hypoth-esized antedating status of the verbal and performance disparity in producing later antisocialbehavior. It can therefore be argued that the cognitive imbalance–antisocial behavior formula-tion constitutes a causal model that could be evaluated via structural equation modeling pro-cedures.

The espoused purpose of structural equation modeling (SEM) is the testing of causal the-ories with nonexperimental data (Martin, 1987). The investigator must first specify and expli-cate a model involving covariance between and among variables. This is in contrast to traditionalstatistical methods such as analysis of variance and multiple regression that focus on models ofindividual observations (Hoyle & Smith, 1994). In using structural equation procedures, it isnoteworthy that relationships encompassing both shared and unique variance can be evaluated(see Hoyle & Smith, for further discussion). If the difference between higher performance IQand lower verbal IQ is critical in relation to antisocial inclinations (assuming the veracity ofintelligence imbalance hypothesis), this would translate into a structural equation path involv-ing the unique variance of performance intelligence; that is, variance that is not shared withverbal IQ. The shared variance between these variables is not an ostensive part of the imbal-ance theorizing. But if, in theory, the unshared variance in performance IQ corresponds toimpulsivity, perhaps the shared component involves the inductive, deductive, and other reason-ing processes common to verbal and performance subtests of intelligence. If so, this sharedvariance might be a latent variable that is the antithesis of the unique performance variance,viz. a variable that has implications for the rational control of impulses. In the parlance ofJoseph’s (1982) theorizing, the shared variance might be viewed as the more mature aspect ofcognitive functioning.

The current investigation was initiated with the following aims: to apply structural equa-tion modeling procedures for the first time in the investigations of imbalance–antisocial behav-ior formulation and to evaluate the robustness of cognitive disparity theorizing by extendingthe research to adult participants.

METHOD

Subjects and Procedure

The participants were 325 adults (120 female: 225 male) ranging in age from 17 to 56. Of thisnumber 141 of the participants were prison inmates; the remainder was composed of individ-uals evaluated by the state vocational rehabilitation services. They were administered the Wech-sler Adult Intelligence Scale-Revised (WAIS-R) and the Minnesota Multiphasic PersonalityInventory-2 (MMPI-2). Corresponding to the model being tested, (see Figure 1) the Verbal andPerformance scales of the WAIS-R and the Psychopathic Deviate (Pd) and Mania (Ma) scoresfrom the MMPI-2 were included as manifest (directly measured) variables. The Pd and Mascores, in combination, are presumed to measure characteristics consistent with antisocial per-sonality disorder (see Graham, 1987). Structural equation modeling procedures of EzPATH(Steiger, 1989) were then applied to the data.

352 Journal of Clinical Psychology, June 1997

RESULTS AND DISCUSSION

A comparison model was initially tested. It was identical to the depiction in Figure 1 except forthe deletion of unique performance IQ variance (again, the variance that was not shared withverbal intelligence). This was done because unique performance variance was construed as thekey element in cognitive imbalance theorizing. The comparison model was found to provide apoor fit to the data (m 5 98.59,p , .00001). A modification of the model consistent with theguiding hypothesis, added unique performance IQ variance and included a fixed (1.0) path tothe latent variable termed “antisocial proclivities” as shown in Figure 1. An insignificant ChiSquare value of 2.49 (p , .11), indicated that an adequate model fit was possible. Indeed, theJöreskog-Sörbom Goodness of Fit Index (GFI) yielded a value of .9960, reflecting an excellentcongruence between model and data. Moreover, the Jöreskog-Sörbom Adjusted Goodness ofFit Index (GFI adjusted for degrees of freedom) of .960 and the Steiger-Lind residuals rootmean square of .088 (average of residual correlations) supported an obtained well-fitted model.Finally, the Adjusted Population Gamma Index, the coefficient of model determination adjustedfor model complexity, was .9612, indicating an excellent convergence of model specificationsand obtained data. Thus, all statistical indices support the adequacy of the hypothesized model.Further, the significant negative path coefficient from the rational control construct to theantisocial variable fortifies the notion that shared verbal and performance variance relates toreasoning processes while unique performance variance reflects impulsivity.

Our findings corroborate Joseph’s (1982) theorizing and support the utility of the devel-opmental arrest, cognitive disparity theorizing with an adult population in comparison to earlierresearch with younger individuals. Furthermore, this support was obtained using the proce-

Figure 1. Cognitive imbalance model of antisocial proclivities.m 5 2.49,p 5 .11. Goodness of fit index5 .996:adjusted goodness of fit index5 .960: root mean square5 .088: adjusted gamma index5 .9612. All standardizedfactor loadings and the standardized regression coefficient are significant atp , .001.

Imbalance and Antisocial Personality 353

dures of structural equation modeling. The statistical tests of SEM verify the theoretical link-ages among constructs involving brain functioning, impulsively, and antisocial proclivities andfurther suggest possible causal status for the cognitive imbalance component of the model.

REFERENCES

Cornell, D., & Wilson, L. (1992). The PIQ.PIQ discrepancy in violent and nonviolent delinquents.Journal of Clinical Psychology, 48,256–261.

Graham, J.R. (1987).The MMPI: A practical guide.New York: Oxford University Press.

Hoyle, R.H., & Smith, G.T. (1994). Formulating clinical research hypotheses as structural equationmodels: A conceptual overview.Journal of Consulting and Clinical Psychology, 62,429–440.

Joseph, R. (1982). The neuropsychology of development: Hemispheric laterality, limbic language, andthe origin of thought.Journal of Clinical Psychology, 38,4–33.

Martin, J.A. (1987). Structural equation modeling: A guide for the perplexed.Child Development, 58,33–37.

Miller, A. (1987). Neurology of the aggressive psychopath: An integrative review.Aggressive Behavior,13, 119–140.

Petee, T.A., & Walsh, A. (1987). Violent delinquency, race, and the Wechsler performance-verbaldiscrepancy.Journal of Social Psychology, 127,353–354.

Steiger, J.H. (1989).EzPATH: Causal modeling.Evanston, IL: Systat, Inc.

Walsh A. (1991).Intellectual imbalance, love deprivation and violent delinquency: A biosocial perspec-tive. Springfield, IL: Charles C. Thomas.

354 Journal of Clinical Psychology, June 1997