Correctional Service Canada
Symbol of the Government of Canada

Common menu bar links

FORUM on Corrections Research

Warning This Web page has been archived on the Web.

Offender Typologies: Identifying Treatment-Relevant Personality Characteristics

It is only recently that corrections has recognized the importance of active intervention with offenders to reduce the likelihood that they will return to their former criminal patterns after release from custody. In large part, this recognition has been fuelled by a virtual explosion of research demonstrating the relationship of criminal sentiments, attitudinal factors, belief and value systems and personality to criminal behaviour.(1) These findings have been coupled with research underscoring the effectiveness of treatment programming in changing criminogenic factors and reducing recidivism.(2) As well, research has emphasized the need to identify treatment-relevant offender characteristics, or typologies, and to match treatment with these characteristics.(3)

The present study was designed to evaluate the ability of a personality assessment instrument to divide offenders into treatment-relevant groups. We used the Millon Clinical Multiaxial Inventory (MCMI) - a 175-item, paper-and-pencil self-report questionnaire designed to measure psychopathology.(4) The MCMI pro-duces a personality profile based on 20 personality dimensions.

The study sample comprised 135 offenders who completed the MCMI within three weeks of arriving at a medium-security federal institution. The MCMI was part of a larger assessment battery which also looked at I.Q., anxiety, depression and socially desirable responding. Data on offender age, ethnic background and criminal history were also collected.

Results

Offenders in the sample averaged about 28 years old, with a range of 19 to 55 years. About 59% were Caucasian, about 38% were native, and 3% were from other ethnic backgrounds.

Four percent of the sample were serving life sentences. For the rest, the average sentence length was 3.8 years, with a range from 2 to 1l years. Despite the fact that the vast majority (83%) were serving their first federal term of incarceration, on average, offenders in our sample had been convicted of almost 22 individual offences over the course of their criminal careers, with a range from 1 to 77 offences.

Cluster analysis was used to classify offenders on the basis of their scores on the MCMI dimensions. This statistical technique, which attempts to group individuals according to shared characteristics, yielded five offender types (see table). It should be noted that these offender types or typologies were not based on observed behaviour, nor were they intended to predict behaviour.


Table 1
Average * MCMI Scores for Each of the Five Groups **
Subscale
Group
1
2
3
4
5
Schizoid-Associative
67.8
83.22
53.3
21.3
45.4
Avoidant
82.9
81.3
59.0
18.7
35.0
Dependent-Submissive
84.6
83.3
61.7
42.4
55.6
Historic-Gregarious
69.3
29.6
72.0
76.0
53.9
Narcissistic
65.2
40.2
81.0
86.9
61.7
Antisocial-Aggressive
59.2
45.7
76.3
76.4
52.0
Complusive
30.1
58.8
42.8
60.9
71.2
Passive-Aggressive
90.9
66.5
67.1
34.5
23.0
Schizotypical
63.0
62.8
53.4
26.1
46.8
Borderline (Personality)
72.9
59.2
58.4
35.5
46.8
Paranoid
70.4
59.8
79.6
66.2
56.8
Anxious
89.9
75.2
63.2
31.9
57.5
Somaticizing
73.4
58.9
58.8
36.9
57.9
Hypomatic
76.6
33.2
69.9
55.4
17.0
Dysthymanic
82.6
74.1
58.1
28.0
60.4
Alcohol Abusive
85.3
79.6
68.2
60.5
45.1
Drug Abusive
89.0
63.1
84.4
78.8
50.6
Psychotic Thinking
68.0
66.0
62.4
45.9
48.4
Psychotic Depressive
67.6
61.8
51.8
38.4
41.5
Psychotic Delusional
63.7
69.1
69.1
58.2
55.8
* Means were calculated for the average.
** Noteworthy subscale scores appear in boldface.

Group 1 represented 20% of the sample and consisted of offenders who experienced significant levels of anxiety and depression. They were likely to be subservient and dependent individuals who were non-assertive and non-confrontational, and who actively avoided interpersonal conflict. As a result of their submissiveness, they tended to ruminate over feelings of anger, resentment and hostility, and expressed these strong emotional reactions through indirect, or passiveaggressive, means. These offenders had the potential for serious drug and alcohol abuse.

Group 2 represented 10% of the sample and consisted of offenders who were socially withdrawn and interpersonally isolated and anxious. Low energy levels, apathy, complacency, helplessness and superficial compliance were also characteristics of these offenders, as was the potential for alcohol abuse.

Egoism and narcissism dominated the personality of offenders in Group 3, who represented about 20% of the sample. They were the youngest group of offenders, with an average age of just over 25 years. Significant anger, aggression and impulse control problems were typical for offenders in this group. Distrust, suspicion and noncompliance were also dominant features. These offenders were likely to demonstrate antisocial attitudes and values, and they were at risk for drug abuse.

Group 4 offenders, representing 24% of the sample, were manipulative smooth talkers, who were likely to seem gregarious and outgoing. However, their veneer of sociability easily gave way and they became confrontational, argumentative and aggressive when their manipulative strategies failed to pay off. In short, these offenders were emotionally volatile and experienced difficulty with anger and impulse control. They were also likely to have drug problems.

The final group comprised 26% of the sample. Offenders in Group 5 lacked the high scores on individual subscales evident in the other profiles. They were the oldest group of offenders, with an average age of almost 31 years. On the surface, these offenders appeared to be the least pathological of the five groups, with no significant concerns regarding substance abuse. They were likely to be friendly, co-operative and socially appropriate. However, they also tended to be somewhat rigid and overly compliant. Indeed, results suggest that they may have been actively trying to foster an unrealistically positive impression, while, at the same time, exhibiting significant levels of denial and self-deception. As a result, this response style may have influenced the profile for this group.

Finally, none of the groups differed with respect to I.Q. or ethnic composition. Despite the fact that the groups did not differ with respect to offence type or the proportion of violent offenders, groups 2 and 4 tended to have more extensive criminal histories, and Group 1 offenders had the greatest number of offences on their present conviction.

Discussion

The results of our analysis indicate that the personality structures of offenders differ dramatically from one another, and that it is possible to group offenders into meaningful personality typologies. Moreover, we feel that our analysis has important implications for treatment.

Although the overall aim of clinical intervention applied within a correctional setting is to facilitate offenders' successful reintegration into society, choice of treatment regime may enhance the effectiveness of therapeutic efforts.

A brief analysis of the five groups reveals a number of salient issues for treatment. To begin with, the combination of personality characteristics displayed in these groups suggests that, in general, these offenders present a challenge for the practitioner. In some instances, they may be reluctant to become involved in therapy due to their avoidance orientation (e.g., Group 1), their apathy (e.g., Group 2) or their suspicion (e.g., Group 3); in other instances, they may reject clinical interpretations because they do not perceive their behaviour to be genuinely problematic (e.g., Group 3 and Group 4).

The manipulative style of Group 4 offenders poses a unique challenge for the therapist particularly when coupled with their confrontational, aggressive and antagonistic orientation. In short, the practitioner must maintain control in therapy.

Group 1, Group 3 and Group 4 offenders experienced significant anger problems. Whereas the development of anger awareness and control may be the primary objective for offenders in groups 3 and 4, the passive-aggressive nature of the reaction of Group i offenders to interpersonal conflict suggests that these individuals need to develop more appropriate methods of expressing extreme emotions.

Group 1 and Group 2 offenders would benefit from more general assertiveness training as well as from treatment aimed at increasing selfefficacy and self-worth. In addition, pharmacological intervention may be warranted for some Group 1 offenders given their high levels of chronic anxiety and depression. However, this latter treatment alternative must be undertaken with caution due to the high risk these offenders present for substance abuse and addiction.

Group 5 offenders were a somewhat unique group of individuals. In addition to being the oldest, their MCMI profile indicated that they were not psychopathological. However, it is likely that their extensive use of impression management, denial and seif-deception suggests that they attempt to present themselves in the best possible light, minimizing or denying less positive personality characteristics. These findings are consistent with recent research(5) which demonstrates that, compared with younger offenders, older offenders are more likely to use the response style of impression management when responding to personality instruments in order to present themselves favourably. Breaking through this barrier of denial and minimization is an important goal for therapy.

In summary, our findings suggest that the MCMI, a standardized self-report instrument, yields useful information regarding the differential personality make-up of offenders.

As a final note, we feel that the high level of offending evidenced in our sample, coupled with serious psychopathology in most groups, underscores the need for effective programming and treatment interventions in correctional settings if we are to reduce recidivism among offenders.


(1)D.A. Andrews, I. Zinger, R.D. Hoge, J. Bonta, P. Gendreau and F.T. Cullen, "Does Correctional Treatment Work? A Clinically Relevant and Psychologically Informed Meta-Analysis," Criminology, 28(1990): 369-404.
(2)
D.A. Andrews, J. Bonta and R.D. Hoge, "Classification For Effective Rehabilitation: Rediscovering Psychology," Criminal Justice and Behavior, 17(1990): 19-52.
(3)
Ibid. See also H.M. Annis and D. Chan, "The Differential Treatment Model: Empirical Evidence From a Personality Typology of Adult Offenders," Criminal Justice and Behavior, 10 (1983): 159-173. And see F.B. Glaser, "The 'Average Package of Help Versus the Matching Hypothesis: A Doggerel Dialogue," Journal of Studies on Alcohol, 38(1977): 1819-1827.
(4)
T. Millon, The Millon Clinical Multiaxial Inventory (Minneapolis, Minn.: NCS, 1983).
(5)D.G. Kroner and J.R. Weekes, "Response Style Measures in an Inmate Sample." Manuscript submitted for publication, 1992.