FORUM on Corrections Research

The utility of Interpersonal Circumplex Theory in research and treatment of sexual offenders

Doctoral Thesis, Queen’s University¹
Dana Anderson²
Advisor: W. L. Marshall
Committee Members: Fred Boland, and Lee Fabrigar

Much of the research on sexual offenders has focused on differences between sexual offenders and non-sexual offenders on a variety of psychological variables (such as, empathy, self-esteem, intimacy deficits, and loneliness). Clinicians and researchers have also noted differences between different types of sex offenders (for example, rapists and child molesters) on these variables. Some of these variables may be linked to problematic interpersonal styles of the offenders. The purpose of this study was to examine differences in interpersonal style between different types of sex offenders and non-sexual offenders using Interpersonal Circumplex Theory as the theoretical framework guiding the investigation.

Interpersonal Circumplex Theory and relevance to sex offenders

Leary³ summarized extensive work by himself and his colleagues on the Interpersonal Circle. This model depicts a method of classifying interpersonal behaviours in a circular structure where the orthogonal axes represent dimensions of control and affiliation. Each of these dimensions has as its anchors bipolar adjectives descriptive of the extremes of these concepts: the control dimension is characterized by dominance or submission, and the affiliative dimension is characterized by friendliness or hostility. These four interpersonal styles comprise the following quadrants of the Interpersonal Circle: hostile-dominant, hostile-submissive, friendly-dominant, and friendly-submissive.

Avector of the Interpersonal Circle is created by the angle from the origin, and this represents one’s interpersonal style. The distance of that vector from the origin is representative of the extent to which that person exhibits that interpersonal style. Therefore, there are different levels along a continuum of interpersonal behaviour which are represented in the model. The more extreme the manifested behaviour, the more rigid the individual is said to be in his or her interpersonal style. Extreme rigidity is associated with severe interpersonal difficulties and psycho-pathology.⁴ For example, an individual may display interpersonal behaviour that is mapped along the vector halfway between dominant and hostile; this behaviour would be classified as “mistrusting.” However, the second, more extreme, level of this behaviour is labelled “suspicious-resentful” and the most extreme level of this behaviour (and hence the level most associated with psychopathological disturbance) is “paranoid-vindictive”.⁵

Some research conducted on the maladjusted interpersonal behaviours of sexual offenders suggests that the interpersonal circumplex model may provide a useful framework for integrating existing theories. Sexual offenders have been shown to be less inclined to exhibit empathy than nonoffenders,⁶ and difficulties with empathy could be related to problematic interpersonal styles. Moreover, sexual offenders have demonstrated intimacy deficits and present as more lonely than non-offenders or non-violent offenders.⁷ Again, these are interpersonal difficulties that can stem from maladaptive interpersonal styles. Because there have been differences in these constructs between different types of sexual offenders,⁸ it follows that different types of sexual offenders would also exhibit different interpersonal styles.

Methodology

The study assessed 148 offenders from a variety of penitentiaries representing all security levels in the Ontario region. They were broken down into the following offender groups based on criminal history: 39 rapists, 42 child molesters, 41 violent (non-sexual) offenders, and 26 general offenders (non-sexual, non-violent). Their average age was 37.42 years (SD = 10.98; range 20-69). Forty-six participants were serving an indefinite sentence (including 14 dangerous offenders). The average sentence length for the remaining 102 men serving definite sentences was 6.43 years. None of the sexual offenders had completed sex offender programming.

The assessment of interpersonal ratings was the Revised Interpersonal Adjectives Scale (IAS-R).⁹ This measure is a 64-item, self-report measure of interpersonal traits. The inventory requires respondents to rate themselves according to the descriptive accuracy of each adjective on an 8-point scale, ranging from “Extremely inaccurate” to “Extremely accurate”. The scale provides scores for the participants on each of the following 8 categories of interpersonal style: Assured-Dominant, Gregarious-Extraverted, Warm-Agreeable, Unassuming-Ingenuous, Unassured-Submissive, Aloof-Introverted, Cold-Hearted, and Arrogant-Calculating. In addition, the scoring program for the IAS-R provides interpersonal profiles. For the purpose of the study, the mean responses to each item for each offender group were inputted to obtain the profile for that group of respondents.

Results

An analysis of variance was computed comparing mean ratings on each octant (or each of 8 interpersonal categories) of the IAS-R for the sex offenders. The differences in mean ratings were statistically significant (p < .05) for five of the categories. Rapists rated themselves, on average, significantly higher than child molesters on traits falling into the following categories: Assured-Dominant, Arrogant-Calculating, Cold-Hearted, and Aloof-Introverted; and they rated themselves significantly lower than child molesters on traits describing a Warm-Agreeable interpersonal style. When all offender groups were included in the analyses, significant differences emerged for the following categories: Arrogant-Calculating, Cold-Hearted, and Aloof-Introverted. The rapists had the highest mean scores of all offender groups for each of those three categories.

The interpersonal profiles were computed for each group. Each profile represents a characteristic configuration for that group. The Angular location indicates the interpersonal type category, and the vector length indicates the intensity with which the behaviours are manifested. It is worth noting that the variables, or interpersonal adjectives comprising the circumplex, are distributed continuously, and individuals falling within the same category may differ from each other. For example, the adjective “domineering” falls within the octant of “assured-dominant” but so does the adjective, “self-assured”. So the variables within each category go along a continuum and the closer they are to the next category, the more closely associated they would be with those adjectives.

The average profile for the rapists indicates the type described as “cold-hearted”. Individuals classified in this sector tend to deny being warm, kind, or sympathetic. In its most problematic form, individuals exhibiting this style have difficulty expressing affection toward others, getting along with others, and forgiving others. In contrast, the profile for the child molesters indicates the type described as “unassured-submissive”. Individuals classified in this sector tend to be timid and submissive in social transactions, lack confidence, and are low in self-esteem. As this style becomes more problematic, such individuals have difficulty making their needs known to others and difficulties with being assertive. The average profile for the violent offenders indicates the type described as “cold-hearted” (described above), whereas the profile for the non-sexual, non-violent offenders indicates the type described as “warm-agreeable”. Individuals in this sector tend to be nurturant, sympathetic, and caring in social transactions. In its problematic form, this interpersonal style is associated with trying too hard to please others, and being too trusting and permissive in dealing with others.

Discussion

Overall, it has been demonstrated that the interpersonal style of rapists differs from that of child molesters, and the interpersonal style of rapists appears to be similar to that of other violent, non-sexual offenders. Those without offences against people (for example, the general offenders) were more likely to endorse adjectives indicative of a more adaptive interpersonal style. It is worth noting that in studies of personality disorders, the “warm-agreeable” octant was unrelated to any of the personality disorders of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV). An interesting point is that the IAS-R is a self-report measure. Therefore, the offenders in this study described themselves this way. This suggests that they do seem to have a conception of their interpersonal style that coincides with a clinician’s impression and with other research investigating some of the differences between child molesters and rapists.

Given how each group rates itself, it is important to consider the hypothesis of Interpersonal Circumplex Theory that different interpersonal styles elicit certain interpersonal behaviours in reaction and keep a pattern going. Ideally, to function effectively on an interpersonal level, an individual should have a repertoire of interpersonal behaviours from which he or she chooses to display the most appropriate behaviour for the given situation.¹⁰ An individual who is fixed on a particular interpersonal style must first be able to modify the cluster of behaviours that define the style, so that the interpersonal style becomes less extreme, and then he or she can more easily move out toward another continuum of behaviour. Therefore, one goal of the therapist should be to coax the client’s behaviour away from its usual extreme. That is, the therapist cannot be lured into displaying complementary behaviour to the client, which would reinforce the client’s already troublesome style.

While some previous researchers and clinicians have suggested certain ideal characteristics of therapists for sex offender groups,¹¹ these findings suggest a need for flexibility in the interpersonal style of the therapists involved in sex offender treatment. In other words, to effectively deal with clients of various interpersonal styles, this theory suggests that a therapist should be able to exhibit interpersonal behaviour that ranges along both dimensions of control (such as, from dominant to submissive) and affiliation (for example, from hostile to friendly). Although it may sound like a strange, if not unethical, suggestion that a therapist display a style that could be considered hostile, the findings of differences in the maladaptive styles of child molesters and rapists do suggest that certain interpersonal styles of therapists may be a better “fit” than others with different types of offenders if one goal is to help an offender to modify his behaviour, rather than being baited by his interpersonal style and inadvertently reinforcing it.

In conclusion, Interpersonal Circumplex Theory provides an interesting theoretical framework for investigating the difficulties observed in many of our clients. Furthermore, the theory provides testable hypotheses to conduct such investigations that may not only provide empirical evidence for clinical impressions, but also to provoke us to consider a variety of methods of treating sex offenders.

---

1.  Abstract from Anderson D. (2001). The utility of Interpersonal Circumplex Theory in research and treatment of sexual offenders. Doctoral Thesis. Kingston, ON: Queen’s University.

2.  340 Laurier Avenue West, Ottawa, Ontario, K1A 0P9.

3.  Leary, T. (1957). Interpersonal diagnosis of personality. New York, NY: Ronald.

4.  Wiggins, J. S., Trapnell, P., and Phillips, N. (1988). Psychometric and geometric characteristics of the revised Interpersonal Adjective Scales (IAS-R).Multivariate Behavioral Research, 23, 517-530.

5.  Kiesler, D. J. (1983). The 1982 interpersonal circle: Ataxonomy for complementarity in human transactions. Psychological Review, 90, 185-214.

6.  Williams, L. M., and Finkelhor, D. (1990). The characteristics of incestuous fathers: Areview of recent studies. In W. L. Marshall, D. R. Laws, & H. E. Barbaree (Eds.), Handbook of Sexual Assault: Issues, Theories, and Treatment of the Offender, New York, NY: Plenum Press, 231-255.

7.  Seidman, B. T., Marshall, W. L., Hudson, S., and Robertson, P. J. (1994). An examination of intimacy and loneliness in sex offenders. Journal of Interpersonal Violence, 9, 518-534.

8.  Garlick, Y., Marshall, W. L., and Thornton, D. (1996). Intimacy deficits and attribution of blame among sexual offenders. Legal and Criminological Psychology, 1, 251-258. See also Seidman, Marshall, Hudson, and Robertson (1994).

9.  Wiggins, J. S. (1995). Interpersonal adjective scales: Professional manual. Odessa, FL: Psychological Assessment Resource.

10.  Knapp, M. L., and Vangelisti, A. L. (1992). Interpersonal communication and human relationships. Needham Heights: Allyn and Bacon.

11.  Marshall, W. L., Anderson, D., and Fernandez, Y. (1999). Cognitive-behavioral therapy with sexual offenders. London, UK: John Wiley and Sons.