Conceptualization and Measurement of Criminal Thinking

Abstract

This article describes two studies concerning the development of a new measure of criminal thinking, the CriminogenicThinking Profile (CTP), influenced by the construct of psychopathy, and traditional models of cognitive-behavioral therapy (CBT). An experimental item pool based on verbalizations from offenders served as the pilot version of the instrument. Principal components analysis of the items resulted in a 62-item, eight-factor scale that was internally consistent. In terms of content, six of the resulting factors were conceptually related to psychopathy, one to CBT, and one to neutralization theory. The factor structure and internal reliability was supported by a subsequent confirmatory factor analysis. Initial support for the CTP’s convergent validity was indicated by its positive correlations with psychopathy and personality disorders associated with criminal, aggressive, and impulsive behaviors. The CTP’s divergent validity was supported by its inverse correlations with indices of healthy personality functioning. The CTP offers a somewhat different constellation of thinking patterns than those found on previously published criminal thinking instruments. The utility of the CTP to identify relevant cognitive targets for offender treatment is a recommended area of future research.

Keywords

criminal thinking offender assessment CTP

One important aspect of the cognitive revolution in the social sciences has been the development of cognitive-behavioral therapies (CBT) that have established themselves as our most empirically supported treatments for a wide variety of disorders (for a recent meta-analytic review, see Butler, Chapman, Forman, & Beck, 2006). Although CBT began with the treatment of depression and anxiety (Beck, 1963, 1967; Ellis, 1957, 1962), it has evolved to include a diverse array of problems, including criminal behavior (Beck, Freeman, Davis, & Associates, 2004). The identification of specific thinking patterns associated with problematic emotional reactions and dysfunctional behaviors is at the core of CBT interventions. One challenge for cognitive-behavioral assessment and treatment in forensic settings is the conceptualization and identification of thinking patterns typical for offenders.

The identification of offender-relevant thinking patterns is critical if CBT interventions provided in forensic settings are to match the effectiveness achieved in programs delivered in traditional mental health settings. Beck’s (1967) and Ellis’s (1962) conceptualizations such as awfulizing (exaggerating the consequences or level of hardship associated with difficult or challenging situations), fortune telling (predicting the future negatively), and personalizing (attributing a disproportionate amount of blame to oneself rather than considering other factors) may not be as broadly applicable to offender-clients as they are for those suffering from traditional mental health disorders. For example, chronic offenders do not seem prone to harshly criticize or blame themselves when they receive negative feedback, as Beck would suggest is the pattern for depressives. Nor is their problem that of awfulizing or overestimating negative outcomes, as Ellis proposes is common for those suffering from anxiety. Criminal thinking patterns may be just the opposite, showing a tendency to minimize responsibility for negative outcomes and to underestimate danger and risk in favor of overly optimistic and self-serving predictions (Maruna & Mann, 2006).

Although the traditional cognitive targets of CBT may be limited in their usefulness for offenders, attempts to conceptualize more relevant thinking patterns have emerged from sources such as Sykes and Matza’s (1957) neutralization theory, Yochelson and Samenow’s (1976) book The Criminal Personality, and Mylonas and Reckless’s (1963) research on offender attitudes toward the criminal justice system. This literature on offender attitudes has formed the foundation for criminal thinking instruments that have been developed or revised over the past 15 years. These include the Psychological Inventory of Criminal Thinking Styles (PICTS; Walters, 1995), Criminal Sentiments Scale–Modified (CSS-M; Simourd, 1997), Measure of Criminal Attitudes and Associates (MCAA; Mills, Kroner, & Forth, 2002), Texas Christian University Criminal Thinking Scales (TCU CTS; Knight, Garner, Simpson, Morey, & Flynn, 2006), and the Measure of Offender Thinking Styles (MOTS; Mandracchia, Morgan, Garos, & Garland, 2007). Each instrument is briefly reviewed below.

The PICTS is an 80-item self-report instrument measuring eight criminal thinking patterns (Walters, 1995) conceptually derived from those identified by Yochelson and Samenow (1976) as well as its author’s clinical experience. Several thinking patterns measured by the PICTS include Cutoff (suppression of thoughts and feelings that would deter criminal activity), Superoptimism (overestimating the likelihood of success related to risky behaviors), and Power Orientation (need to control others and events). Factor analyses have suggested that the items load onto two primary and two minor factors. The primary factors are (a) Problem Avoidance (seeking shortcuts/distractions/avoidant strategies when confronted with problems) and (b) Self-assertion/Deception (justifying criminal behavior and overestimating its likelihood of success). The two minor factors are (a) Interpersonal Hostility (hostile attitude toward others/arrogance) and (b) Denial of Harm (denial and minimization of harm caused by one’s criminal behavior; Walters, 1995, 2005). Walters prolific research program has recently focused on higher order factors related to criminal behavior such as proactive criminality, which is described as planned and goal directed, and reactive criminality, which is unplanned, emotional, and impulsive. In terms of thinking, a proactive style is associated with positive outcome expectancies for crime whereas a reactive style seems more associated with hostile and exaggerated attributions (Walters, 2008; Walters, Felix, & Reinoehl, 2009).

The CSS-M is a 41-item self-report instrument measuring five criminal thinking patterns. Three of the thinking patterns concerning attitudes toward the law, courts, and police combine to form a single subscale, Law-Court-Police (LCP), measuring negative attitudes toward the criminal justice system and are derived from the work of Mylonas and Reckless (1963). One subscale, Tolerance for Law Violations (TLV; justification for criminal behavior) is drawn directly from neutralization theory (Sykes & Matza, 1957), whereas the last subscale, Identification With Criminal Others (ICO; perceived similarity to antisocial peers), seems linked with differential association theory (Sutherland & Cressey, 1978). Factor analytic research suggests the CSS-M is four factors rather than three (Simourd & Olver, 2002), with the LCP subscale being divided into two constructs: one relating to attitudes toward criminal justice personnel and the other focused on attitudes toward the legal system.

The MCAA is a two-part instrument. The first part is unrelated to criminal thinking and quantifies the amount of time an offender spends with criminal peers. The second portion of the scale consists of 46 items measuring four criminal thinking patterns, each supported by a factor analysis (Mills et al., 2002). The first, Entitlement, is influenced by the work of Walters (1995). Antisocial Intent, Violence, and Antisocial Associates are linked to the empirical literature on predictors of recidivism.

The TCU CTS is comprised of 37 items and contains six subscales: Entitlement, Justification, Personal Irresponsibility, Power Orientation, Cold Heartedness, and Criminal Rationalization (Knight et al., 2006). The development of these subscales was influenced by the work of Walters (1995). The TCU CTS appears to have four factors where Entitlement, Justification, and Personal Irresponsibility combine to form one factor, whereas the other three factors correspond with the original subscales.

Finally, the MOTS was based on analyzing 77 thinking patterns culled from the work of Walters, Yochelson and Samenow, Beck, and Ellis (Mandracchia et al., 2007). Each thinking pattern was measured with three items that were averaged to produce a score for the pattern. Rather than factor analyze the individual scale items, the thinking pattern scores were the focus of the authors’ analyses. A three-factor model emerged as the most interpretable: Control (need for power control over self, others, and environment), Cognitive Immaturity (lazy and undisciplined decision making and problem solving), and Egocentrism (self-centered and entitled).

Notably lacking from existing criminal thinking instruments is the influence of the personality construct of psychopathy. Characterized by egocentricity, lack of empathy and remorse, a parasitic and exploitive orientation toward others, thrill seeking, and irresponsibility (Hare, 1996), psychopathy has been found to be clinically challenging to treat (Thornton & Blud, 2007) and a predictor of recidivism (Wormith, Olver, Stevenson, & Girard, 2007), violence (Campbell, French, & Gendreau, 2009), and institutional misconduct (Edens, Poythress, Lilienfeld, & Patrick, 2008). Although descriptions of a personality style similar to psychopathy can be traced back to hundreds of years, modern conceptualizations are based on Cleckley’s (1964) book The Mask of Sanity and Hare’s (2003) operationalization of the construct with the Psychopathy Checklist–Revised (PCL-R). Psychopathy appears to be comprised of two correlated factors. Factor 1 encompasses traits such as egocentricity, remorselessness, and manipulativeness, which mark the psychopath’s interpersonal style, whereas Factor 2 includes the antisocial, impulsive, and unstable lifestyle of the psychopath (Harpur, Hare, & Hakstian, 1989). A number of key controversies, such as symptom stability, the potential treatability of individuals who meet the criteria for the syndrome (Edens, 2006), and whether psychopathy represents an extreme variant of normal behavior or a discrete class of people (Wright, 2009), remain unresolved. Most relevant for the focus of this article, several authors have noted the relative lack of assessment around the cognitive component of psychopathy that may serve to drive criminal behavior (Gonsalves, Scalora, & Huss, 2009).

A literature exploring the relationship between criminal thinking and psychopathic characteristics has begun to emerge. Investigations utilizing the PICTS (Gonsalves et al., 2009), CSS-M (Simourd & Hoge, 2000), and TCU CTS (Dembo et al., 2007) have found positive correlations between criminal thinking and measures of psychopathy, although one study did not find a significant relationship between the PICTS and a screening version of the PCL-R (Walters & Mandell, 2007). Overall, criminal thinking has been more strongly correlated with Factor 2 than Factor 1 (Gonsalves et al., 2009; Simourd & Hoge, 2000). In a recent study using the PCL-R and the PICTS conjointly to predict recidivism in an inpatient forensic sample, Gonsalves et al. (2009) found that the superoptimism scale of the PICTS improved the predictive utility of the PCL-R. Given the established relationship between psychopathy and chronic offending and the potential for identifying specific thinking patterns associated with the syndrome, a criminal thinking scale that specifically takes into consideration cognitions likely to accompany psychopathy would be particularly relevant for adding additional insight into attitudes that set the stage for criminal behavior.

The present article describes the development and initial validation of a new measure of criminal thinking, the Criminogenic Thinking Profile (CTP). The rationale for developing a new instrument was to explore the extent to which thinking patterns influenced by the construct of psychopathy, and traditional models of cognitive-behavioral therapy (CBT), would expand the range of empirically identified criminal thinking patterns already recognized in the existing offender attitude literature. This article presents data from two studies. The first focused on item development, factor structure, internal consistency, and divergent validity. The second focused on confirmatory factor analyses (CFAs) as well as internal consistency and divergent and convergent validity. Both studies were approved by university and agency-specific institutional review boards.

Several specific hypotheses were examined in this initial validation phase. First, it was hypothesized that the CTP would have a multidimensional factor structure. This was expected because existing criminal thinking instruments have anywhere from three (CSS-M and MOTS) to four (PICTS, MCAA, and TCU CTS) factors. Second, it was hypothesized that CTP factors would be inversely related to healthy personality traits. The limited published research concerning criminal thinking and healthy personality functioning has found negative correlations between criminal thinking and traits such as agreeableness and conscientiousness (Bulten, Nijman, & van der Staak, 2009; Egan, McMurran, Richardson, & Blair, 2000). Third, it was predicted that the CTP factors would be more strongly associated with personality disorders related to criminal, aggressive, and impulsive behaviors than with personality disorders associated with traditional mental health problems. Although research on criminal thinking and personality disorders is limited, two studies found significant relationships between the PICTS and measures of antisocial personality (Bulten et al., 2009; Walters & Geyer, 2005), whereas one study of delinquent youth found a significant relationship between the TCU CTS and a measure of conduct disorder (Dembo et al., 2007). Fourth, based on prior studies linking criminal thinking to psychopathy, as well as the influence of psychopathy on the development of the CTP, strong correlations between the CTP and a measure of psychopathy were expected.

Study 1

Method

Study 1 began with the creation of an experimental item pool, which was administered to a large sample of offenders. A principal components analysis (PCA) was used to explore the factor structure of the CTP and to shorten the instrument. This was followed by an examination of internal consistency. Finally, divergent validity was explored by correlating the CTP with a measure of healthy personality functioning.

Participants

A total of 1,529 offenders (1,296 probationers and 233 inmates) recruited from probation department waiting rooms and correctional institutions in the northeastern United States consented to participate. However, 209 participants were unable to complete the survey or were deemed to produce invalid data for the following reasons: (a) omitted more than four items on the survey (n = 113), (b) were called to their probation appointments before the survey could be completed and did not return (n = 27), (c) were observed by research staff to complete the survey randomly, without appearing to read the items (n = 64), or (d) consented to participate but were unable to read English and could not complete the survey (n = 5). The final sample (N = 1,320) consisted of 1,101 probationers and 219 inmates. Table 1 provides demographic data on participants in both studies.

Table 1.

Demographic Characteristics of Participants

	Study 1		Study 2
Characteristic	Probationers (n = 1,101)	Inmates (n = 219)	Probationers (n = 365)
Age
M	31.52	33.29	33.31
SD	11.42	10.70	11.19
Gender
Male	71% (n = 785)	41% (n = 90)	73% (n = 265)
Female	23% (n = 249)	59% (n = 129)	19% (n = 69)
Missing	6% (n = 67)	0%	8% (n = 31)
Ethnicity
White	25% (n = 278)	33% (n = 73)	23% (n = 85)
African American	37% (n = 403)	42% (n = 92)	36% (n = 131)
Latino	26% (n = 282)	22% (n = 47)	31% (n = 113)
Asian American	0.5% (n = 4)	2% (n = 4)	<1% (n = 1)
Native American	1.5% (n = 15)	1% (n = 3)	1% (n = 4)
Other	5% (n = 59)	0%	3% (n = 12)
Missing	5% (n = 60)	0%	5% (n = 19)
Education
Less than 12 years	32% (n = 347)	41% (n = 90)	32% (n = 115)
12 years	42% (n = 459)	40% (n = 88)	43% (n = 152)
More than 12 years	25% (n = 274)	16% (n = 34)	24% (n = 84)
Missing	2% (n = 21)	3% (n = 7)	4% (n = 14)
Arrests
M	5.99	5.07	5.15
SD	7.96	5.80	6.22

Materials

CTP

Items for the CTP were initially generated during training sessions for probation officers and criminal justice case managers on the topic of criminal thinking. Broad categories of thinking styles referenced from CBT (e.g., low frustration tolerance, overgeneralization), and the criminal thinking literature (e.g., underestimating risk, neutralization), were discussed, as well as the protypical characteristics of psychopathy (e.g., lack of empathy, impulsive decision making) characterized by Cleckley (1964) and Hare (1996). The criminal justice practitioners were asked to provide written examples of actual client verbalizations that reflected the thinking styles as well as client verbalizations that did not fit into the preconceived styles but which they believed typified criminal thinking. For example, when lack of concern for others was discussed as a symptom of psychopathy, practitioners were asked to write down recent statements from their clients that reflected this symptom. These verbalizations were collected, reviewed, and edited slightly. Statements reflecting behaviors, rather than thinking or attitudes, were excluded. Extreme profanity was removed and gender-specific language was replaced with gender-neutral language. The resulting verbalizations became the item pool (n = 134) for a pilot version of the CTP. A 4-point Likert-type scale (1= strongly disagree, 2 = disagree, 3 = agree, 4 = strongly agree) followed each item. This somewhat novel approach to item development was adopted because we believe that actual verbalizations by offenders can better reflect the language, content, and manner of thinking than items generated by academics or researchers.

Emotional Quotient Inventory (EQ-i)

The EQ-i (Bar-On, 1997) is a 133-item self-report questionnaire that assesses a broad range of healthy personality traits. The scale yields a total score, 5 composite scale scores (Intrapersonal, Interpersonal, Adaptability, Stress Management, and General Mood), 15 brief subscale scores derived from the five composite scales, and 4 validity scale scores. Respondents rate the degree to which each item describes them on a 5-point Likert-type scale. Higher scores indicate healthier functioning. Psychometric analysis of the EQ-i with offenders has indicated that the scale is internally consistent and that scores on the scale are negatively correlated with emotional problems (Hemmati, Mills, & Kroner, 2004).

Procedure

Probationer participants were recruited from probation department waiting rooms. After probationers checked in at the main desk, they were approached by research staff and asked to complete the CTP, for which they would be compensated US$5. Participation was voluntary, anonymous, and not a condition of probation. Probation officers were not present when the CTP was completed. To obtain a diverse sample in terms of age, gender, ethnicity, and offense backgrounds, probationers were recruited from urban, suburban, and rural probation offices. The number of probationers that refused to participate was not recorded but was estimated at less than 5%.

Inmate participants were recruited from a correctional life skills training program and completed the CTP and the EQ-i on the first day of programming. Inmates received no compensation and the completion of the instruments was voluntary, confidential, and not a condition of the program. Correctional staff did not have access to inmate’s responses. Male inmates were recruited from a medium security institution, whereas female inmates were obtained from an institution comprised of mixed security levels. No inmates declined participation.

Results

PCA

The factor structure of the CTP was examined through a PCA with a varimax rotation. A procedure known as parallel analysis was used to determine the number of factors to be retained (O’Connor, 2000). In parallel analysis, eigenvalues based on a randomized sorting of the data are compared with eigenvalues based on the actual data. A factor is retained if the eigenvalue based on the actual data is greater than the eigenvalue of the randomly ordered data (Thompson, 2004). Parallel analysis has been found to determine the number of factors to retain with greater accuracy than more commonly employed procedures (e.g., eigenvalues greater than 1, scree test) that are more likely to lead to the retention of spurious factors (Patil, McPherson, & Friesner, 2010; Thompson, 2004; Zwick & Velicer, 1986). There were eight factors whose eigenvalues exceeded their randomly ordered counterparts and these became the subscales for the CTP.

Factor items were retained if they loaded >.40 on only a single factor. Items whose removal increased or had no significant influence on the factor’s internal consistency were deleted. This led to a 62-item scale (α = .96). A description of each subscale and corresponding alpha coefficients are presented in Table 2. Alpha coefficients for the subscales ranged from .73 to .92 and show consistency across offenders samples from Study 1 and 2.

Table 2.

CTP Subscale Descriptions and Alpha Coefficients

CTP subscale	Description/sample item	α (Study 1)	α (Study 2)
Disregard for Others (14 items)	Lack of concern, empathy, and/or remorse	.92	.91
	“It doesn’t make sense to feel bad about other people’s problems”
Demand for Excitement (9 items)	Low tolerance for boredom, impulsivity	.87	.90
	“There is no better feeling than the rush I get when stealing”
Poor Judgment (9 items)	Underestimating negative outcomes related to risky behaviors	.88	.89
	“Gambling or playing the lottery is a good investment because if you play long enough you are bound to win”
Emotionally Disengaged (6 items)	Lack of trust, avoidance of emotions that lead to vulnerability and intimacy	.84	.79
	“It is important not to show weakness by letting others know your feelings”
Parasitic/Exploitive (4 items)	Parasitic and exploitive worldview, evading responsibility	.73	.77
	“It doesn’t make sense to work full-time if you can get on a government program”
Justifying (6 items)	Normalization/ minimization of antisocial/self-destructive behavior	.87	.82
	“Someone is going to profit from illegal activity so it might as well be me”
Inability to Cope (7 items)	Giving up in the face of adversity, ineffective problem solving	.83	.82
	“When I don’t understand things I give up”
Grandiosity (7 items)	Overestimating one’s skills, abilities, or inner qualities	.79	.77
	“I have more positive qualities than most people”

Note: CTP = Criminogenic Thinking Profile.

CTP Intercorrelations and Descriptive Data

Intercorrelations among the CTP Total Score and subscales are presented in Table 3. All correlations were statistically significant but not high enough to suggest redundancy in the subscales. The strongest correlation for each subscale was with the total score. With the exception of Grandiosity, intercorrelations among the subscales were moderate to large, ranging from .36 to .76. Correlations between Grandiosity and the other subscales were small to moderate ranging from .15 to .37. Means, standard deviations, and quartile scores are presented in Table 4.

Table 3.

CTP Intercorrelation Matrix (N = 1,320)

Scale	1	2	3	4	5	6	7	8	9
1. CTP Total Score		—
2. Disregard for Others	.91	—
3. Demand for Excitement	.82	.69	—
4. Poor Judgment	.85	.76	.69	—
5. Emotionally Disengaged	.65	.52	.39	.40	—
6. Parasitic/Exploitive	.71	.65	.56	.60	.36	—
7. Justifying	.84	.76	.66	.70	.47	.62	—
8. Inability to Cope	.80	.67	.68	.62	.52	.54	.59	—
9. Grandiosity	.41	.25	.17	.22	.37	.15	.23	.22	—

Note: CTP = Criminogenic Thinking Profile. All coefficients are significant at p < .01.

Table 4.

CTP Means, Standard Deviations, and Quartile Scores (N = 1,320)

Scale	M	SD	25th percentile	75th percentile
CTP Total Score	120.67	28.45	101	138
Disregard for Others	25.86	7.85	19	30
Demand for Excitement	15.18	5.02	11	18
Poor Judgment	15.21	5.20	11	18
Emotionally Disengaged	14.31	3.97	12	17
Parasitic/Exploitive	7.12	2.51	5	9
Justifying	10.89	3.95	7	13
Inability to Cope	13.77	4.27	11	16
Grandiosity	18.34	4.22	16	21

Note: CTP =Criminogenic Thinking Profile.

Correlations Between the CTP and EQ-i

In exploring divergent validity, Table 5 indicates an inverse relationship between the CTP and the EQ-i. The CTP Total Score was moderately to strongly correlated with the EQ-i total score and the five composite scales. The Inability to Cope subscale emerged with the strongest pattern of inverse correlations with the EQ-i. The Grandiosity subscale showed little relationship with the EQ-i, the one exception consisting of a weak positive correlation with the General Mood Composite Scale. Overall, the General Mood Composite Scale of the EQ-i had a weaker pattern of correlations with the CTP than the remaining four composite scores.

Table 5.

Intercorrelations Between CTP and the EQ-i (N = 219)

	EQ-i Scales
CTP scales	Total	Adaptability	Interpersonal	Intrapersonal	Stress	General Mood
CTP Total Score	−.47***	−.48***	−.40***	−.38***	−.35***	−.28***
Disregard for Others	−.40***	−.40***	−.38***	−.32***	−.28***	−.23***
Demand for Excitement	−.44***	−.41***	−.37***	−.36***	−.34***	−.26***
Poor Judgment	−.30***	−.29***	−.29***	−.24***	−.20***	−.14*
Emotionally Disengaged	−.34***	−.32***	−.33***	−.28***	−.25***	−.23***
Parasitic/Exploitive	−.29***	−.30***	−.25***	−.24***	−.23***	−.16*
Justifying	−.36***	−.33***	−.38***	−.30***	−.23***	−.20**
Inability to Cope	−.59***	−.58***	−.43***	−.52***	−.52***	−.40***
Grandiosity	−.04	−.06	.07	.08	−.01	.15*

Note: CTP = Criminogenic Thinking Profile; EQ-i = Emotional Quotient Inventory; Total = EQ-i Total Score; Stress = Stress Management.

p < .05. **p < .01. ***p < .001.

Study 2

Method

To more fully explore the reliability and validity of the CTP, a second round of data collection was conducted. This included analyses of the internal consistency of the shortened instrument and a validation, through CFA, of the factor structure obtained in Study 1. To examine the convergent and divergent validity of the CTP, we explored its relationship with the full range clinical personality disorders and psychopathy.

Participants and Procedure

A total of 434 probationers consented to participate in the study. Data from 69 participants were excluded because they either omitted an excess of four items (n = 27), were observed to respond randomly (n = 8), had language barriers and could not complete the survey (n = 4), or were called to their probation appointments before the survey could be completed (n = 30). The final sample consisted of 365 probationers. Demographic data on the probationers are presented in Table 1. Probationers were recruited in an identical manner as described in Study 1.

In addition to the CTP, subsets of the sample were randomly selected to also complete either the Millon Clinical Multiaxial Inventory–Third Edition (MCMI-III; Millon, Davis, & Millon, 1997) or the Levenson Self-Report Psychopathy Scale (LSRP; Levenson, Kiehl, & Fitzpatrick, 1995). No participant was asked to complete both supplementary measures due to concerns that the added time and length would increase the risk of invalid data or incomplete protocols. Eighty one probationers completed the MCMI-III and 102 completed the LSRP.

Materials

MCMI-III

To assess a range of personality disorders, we chose the MCMI-III (Millon et al., 1997), a 175-item true–false instrument that measures personality and psychopathology and provides scales that closely conform to the Diagnostic and Statistical Manual of Mental Disorders (4th ed.; DSM-IV; American Psychiatric Association, 1994). The MCMI-III provides scores on 10 subscales that represent acute clinical disorders (Axis I) and 14 personality patterns (Axis II). The clinical scales of the MCMI-III have been shown to have good internal consistency and test–retest reliability, and correlate positively with other measures of psychopathology and clinician-judged evaluations (Millon et al., 1997; Strack, 2008). Although data were collected on the full range of disorders measured by the MCMI-III, our analyses are confined to the clinical personality patterns because we were most interested in the hypothesized relationship between the CTP and personality traits.

LSRP

To assess psychopathy, we chose the LSRP (Levenson et al., 1995), a 26-item, self-report, Likert-type instrument. The LSRP yields a total score and two factor scores. Factor 1 is designed to assess signs of psychopathy such as lack of remorse, callousness, and deceitfulness. Factor 2 assesses characteristics of psychopathy such as impulsivity, low frustration tolerance, and poor adherence to social norms. Reports of internal consistency have been adequate for Factor 1 (α = .82) and Factor 2 (α = .63; Levenson et al., 1995), but Factor 2 may be the subscale with greater validity (Lilienfeld & Fowler, 2006). The LSRP correlates with the PCL-R (Brinkley, Schmitt, Smith, & Newman, 2001) as well as thrill seeking and antisocial behavior (Levenson et al., 1995).

Results

Internal Consistency and CFAs

The alpha coefficients for the sample (see Table 2) were similar to those reported for Study 1 and ranged from .77 to .91. The alpha coefficient for the total score was .95. CFA were conducted with EQS for Windows 6.1 (Bentler, 2001). We tested the eight-factor model, and also a one-factor model for comparison purposes, using the maximum likelihood with robust methods procedure because the normalized estimate revealed significant skewness in the data. Model fit was evaluated with multiple indices, principally, the root mean square error of approximation (RMSEA), recommended by Keith (2006) as the best index of fit. As a guideline, models producing RMSEA values less than .1 are considered an adequate fit and values less than .05 are considered a good fit. We also examined the nonnormed fit index (NNFI), normed fit index (NFI), comparative fit index (CFI), and independence fit index (IFI). As a guideline, models producing NNFI, NFI, CFI, and IFI values greater than .90 are considered an adequate fit, and values greater than .95 are considered a good fit.

The results of the CFA are summarized in Table 6. The RMSEA for the eight-factor model was .033, indicating a good fit, whereas the NNFI, CFI, and IFI were above .90, suggesting an adequate fit. Only the NFI fell below .90. Across all the indices, the eight-factor model provided a better fit than the single-factor model. The improvement in fit yielded by the eight-factor model was evaluated by subtracting the chi-square value for the eight-factor model from the chi-square value for the single-factor model, subtracting the degrees of freedom for the eight-factor model from the degrees of freedom for the single-factor model, and evaluating the resulting Δχ² as an ordinary chi-square (Thompson, 2004). This procedure indicated that the eight-factor model provided a significant improvement in fit over the single-factor model, Δχ² = 1373, Δdf = 28, p < .0001.

Table 6.

Summary of Confirmatory Factor Analyses

Model	RMSEA	RMSEA 90% CI	NNFI	NFI	CFI	IFI	Model χ²	df
Eight-factor model	.033	[.030, .036]	.90	.73	.91	.91	2503.16	1801
One-factor model	.055	[.053, .058]	.72	.59	.73	.73	3878.16	1829

Note: RMSEA = Root Mean Square Error of Approximation; CI = confidence interval; NNFI = Bentler–Bonnett Nonnormed Fit Index; NFI = Bentler–Bonnett Normed Fit Index; CFI = Comparative Fit Index; IFI = Bollen’s Independence Fit Index.

Correlations Between the CTP and MCMI-III

Correlations between the CTP and MCMI-III personality disorders are presented in Table 7. The CTP Total Score was most strongly correlated with the aggressive personality disorders (e.g., sadistic, borderline, negativistic, and antisocial), followed by the odd/eccentric personality disorders (e.g., paranoid, schizoid, and schizotypal). In contrast, the CTP Total Score was relatively unrelated to the anxious and depressive personality disorders. The one exception was compulsive personality, which had a significant inverse correlation.

Table 7.

Intercorrelations Between the CTP and the MCMI-III Personality Disorders^a (N = 74)

	MCMI-III personality disorder
CTP scales	Sadistic	Borderline	Negativistic	Antisocial	Paranoid	Schizoid	Schizotypal	Avoidant	Dependent	Depressive	Narcissistic	Compulsive
CTP Total Score	.48**	.39**	.39**	.37**	.29*	.29*	.24*	.22	.18	.10	.05	−.35**
Disregard for Others	.49**	.37**	.28**	.34**	.22	.26*	.22	.20	.17	.16	−.08	−.37**
Demand for Excitement	.35**	.27*	.27*	.37**	.08	.11	.14	.10	.12	−.04	.02	−.26*
Poor Judgment	.25*	.15	.19	.14	.10	.14	.02	.11	.05	−.04	.06	−.26*
Emotionally Disengaged	.12	.20	.25*	.11	.23*	.29*	.14	.13	.03	.08	.10	−.03
Parasitic/Exploitive	.44**	.30**	.41**	.39**	.30**	.25*	.26*	.14	.11	.02	.05	−.28*
Justifying	.44**	.34**	.29*	.36**	.19	.21	.21	.14	.17	.02	.08	−.32**
Inability to Cope	.33**	.34**	.37**	.36**	.25*	.22	.22	.29*	.27*	.21	.16	−.43**
Grandiosity	.06	.02	.05	.16	.27*	.06	.07	.05	−.01	.07	.27*	.16

Note: CTP = Criminogenic Thinking Profile; MCMI-III = Millon Clinical Multiaxial Inventory–Third Edition.

Although the present study concerned the relationship between the CTP and personality disorders, tables containing the correlations between the CTP and all of the MCMI-III clinical scales are available by contacting the first author.

p < .05. **p < .01.

Examination of the relationship between the individual CTP subscales and MCMI personality disorders reveals that Disregard for Others, Demand for Excitement, Parasitic/Exploitive, Justifying, and Inability to Cope subscales were most strongly correlated with the aggressive personality disorders. Poor Judgment was correlated with the sadistic personality but not the other aggressive styles. Emotionally Disengaged was most strongly correlated with the schizoid, negativistic, and paranoid patterns. Not surprisingly, Grandiosity was correlated with the narcissistic personality disorder but equally with paranoid personality. Inability to Cope was the only CTP subscale to correlate with avoidant and dependent, two anxious styles. All of the CTP subscales, except Emotionally Disengaged and Grandiosity, had significant inverse correlations with compulsive personality.

Correlations Between the CTP and LSRP

Correlations between the CTP and LSRP are presented in Table 8. The CTP Total Score was strongly correlated with the LSRP Total Score, supporting a relationship between criminal thinking and psychopathy. Examination of the CTP subscales and the LSRP Total Score reveals a strong correlation with Disregard for Others; moderate correlations with Demand for Excitement, Poor Judgment, Parasitic/Exploitive, Justifying, and Inability to Cope; and a small correlation with Emotionally Disengaged. The only CTP subscale not correlated with the LSRP Total Score was Grandiosity. Examination of the differences between the CTP and the LSRP Factor 1 and Factor 2 subscales reveals that Disregard for Others and Grandiosity had stronger correlations with Factor 1. The CTP Total Score and the Demand for Excitement, Emotionally Disengaged, Parasitic/Exploitive, and Inability to Cope subscales had stronger correlations with Factor 2. Poor Judgment and Justifying correlated equally with both psychopathy factors.

Table 8.

Intercorrelations Between the CTP and LSRP (N = 219)

	LSRP Scale
CTP scales	Total	Factor 1	Factor 2
CTP Total Score	.62**	.47**	.56**
Disregard for Others	.66**	.54**	.47**
Demand for Excitement	.41**	.26**	.41**
Poor Judgment	.41**	.31*	.33**
Emotionally Disengaged	.26*	.18	.39**
Parasitic/Exploitive	.46**	.31**	.45**
Justifying	.45**	.35**	.35**
Inability to Cope	.46**	.21*	.55**
Grandiosity	.03	.22*	−.03

Note: CTP = Criminogenic Thinking Profile; LSRP = Levenson Self-Report Psychopathy Scale; Total = LSRP Total Score.

p < .05. **p < .01.

Discussion

The purpose of this project was to validate a new criminal thinking instrument influenced by the construct of psychopathy and CBT. The development of the CTP differed from other criminal thinking instruments in that statements from offenders served as the basis for item content. This methodology is unique among criminal thinking scales. Our intention in adopting this unorthodox approach to item development was to create an instrument that accurately mirrors the thinking of its intended target population. The CTP yielded more factors than previously published instruments such as the PICTS (Walters, 1995), CSS-M (Simourd, 1997), MCAA (Mills et al., 2002), and TCU CTS (Knight et al., 2006. Of the patterns that emerged on the CTP, six appear readily related to psychopathy (Disregard for Others, Demand for Excitement, Poor Judgment, Emotionally Disengaged, Parasitic/Exploitive, and Grandiosity), whereas Inability to Cope appears analogous to Ellis’s (1962) concept of low frustration tolerance and Justifying appears consistent with Sykes and Matza’s (1957) neutralization theory. Data from two studies provide initial support for the factorial, convergent, and divergent validity of the CTP, and evidence of the instrument’s internal consistency. Results are discussed in terms of the specific hypotheses tested, implications for forensic CBT treatment, and limitations.

Hypotheses

As predicted in the first hypothesis, the present findings are consistent with a multidimensional model of criminal thinking. Study 1 resulted in a 62-item, eight-factor scale with strong internal consistency across the subscales. The subscales’ intercorrelations provide additional evidence of a multidimensional factor structure. Just as one thinking pattern is not likely to be a unifying feature of all mental health problems, a variety of beliefs appear to be associated with criminal behavior. For cognitive-behavioral practitioners working with offender-clients, it seems that a range of thinking patterns could serve as potentially relevant treatment targets.

With respect to the second hypothesis, the CTP was significantly, but inversely, correlated with the EQ-i, a multidimensional measure of healthy personality functioning. With the exception of Grandiosity, CTP subscales were consistently at variance with healthy personality dimensions such as interpersonal skills, adaptability, stress management, self-awareness, and optimism. Inability to Cope, which reflects a passive and ineffective attitude toward facing life’s challenges, was the CTP subscale most strongly and inversely related to the EQ-i’s indices of healthy personality traits. The overall pattern of correlations between CTP and EQ-i subscales provides support for the divergent validity of the CTP and adds to the small body of literature on the relationship between criminal thinking and healthy personality. The present study and those concerning the PICTS and the NEO-Five Factor Inventory Scales of Agreeableness and Conscientiousness have all found that increases in criminal thinking are at odds with those traits generally associated with successful functioning (Bulten et al., 2009; Egan et al., 2000).

As predicted in the third hypothesis, and supporting the convergent validity of the CTP, strong correlations were generally found between the CTP and those personality disorders on the MCMI-III most associated with criminal, aggressive, and impulsive behaviors (sadistic, borderline, negativisitic, and antisocial). The notable exception to our predicted pattern of correlations was the lack of a relationship between the aggressive personality disorders and the Grandiosity subscale. Potential reasons why the Grandiosity subscale did not function in a way that was consistent with the original hypotheses are discussed below. Although there is a paucity of research on criminal thinking and personality disorders, the significant correlation between the CTP and the Antisocial subscale of the MCMI-III is consistent with prior research that has found significant relationships between the PICTS and antisocial personality as measured by the Personality Assessment Inventory (Walters & Geyer, 2005) and the MINI (Bulten et al., 2009).

Also consistent with the third hypothesis, personality disorders least empirically and conceptually related to criminal behavior (depressive, dependent, and avoidant) were not significantly related to the CTP. The one exception was a significant correlation between Inability to Cope and the two anxious personality styles avoidant and dependent. This suggests that offenders who believe that they cannot effectively deal with life’s challenges may also be susceptible to long-term anxiety experiences. The odd/eccentric personality patterns (schizoid, schizotypal, and paranoid), which have been found to be moderately related to criminal behavior, resulted in only a few significant correlations with the CTP. The Parasitic/Exploitive and Emotionally Disengaged subscales seemed to be most connected to the odd/eccentric personality profiles. In addition to difficulties relating emotionally to others, perhaps offenders with schizophrenic spectrum personality styles tend to see themselves as needing other people or institutions to provide for their well-being. Not surprisingly, the CTP Total Score and six of the subscales were inversely correlated with compulsive personality. As compulsiveness is marked by perfectionism, rigid adherence to rules and norms, and hyperconscientiousness, and reflects a personality style that is polar opposite of criminal thinking, this finding lends further support for the divergent validity of the CTP.

Given the influence of psychopathy in the development of CTP items, the fourth hypothesis predicted a strong pattern of correlations between the CTP and psychopathy, as measured by the LSRP. All of the CTP subscales were significantly correlated with one or more LSRP scores. Consistent with research using the PICTS and CSS-M (Gonsalves et al., 2009; Simourd & Hoge, 2000), the CTP was more strongly related to Factor 2 (antisocial behavior/social deviance) than Factor 1 (emotional/interpersonal traits). Inability to Cope was the subscale with the strongest relationship to Factor 2 and suggests that cognitions reinforcing ineffective coping responses are associated with this aspect of the psychopathy construct. Among the other CTP subscales, Emotionally Disengaged, Demand for Excitement, and Parasitic/Exploitive were also more strongly associated with Factor 2 than Factor 1, whereas Disregard for Others and Grandiosity were more strongly related to Factor 1. We see these results as consistent with the Factor 1 and Factor 2 conceptualization of psychopathy. Thinking patterns reflecting poor problem solving, thrill seeking, opportunism and exploitation, and a wariness of relating to others emotionally are more associated with the antisocial, impulsive, and unstable lifestyle aspects of Factor 2. Thinking patterns reflecting lack of remorse and inflated sense of self are more linked with the callousness and egocentricity of Factor 1. Justifying and Poor judgment appeared to be equally related to both factors, suggesting that a tendency to underestimate risky behavior and to minimize or rationalize criminal activity cuts across both psychopathy dimensions.

Limitations

Several limitations exist in these initial validation studies as well as with the criminal thinking perspective in general. First, there were limitations related to the samples. Data on the nature of the offenses that the participants had committed were not obtained. Thus, we were unable to link specific thinking patterns with offense types. In addition, our conservative approach to excluding surveys may have resulted in a sample that was somewhat more organized and compliant than the general population of probationers. A related limitation is the fact that our samples were obtained in a single northeastern state and were English speaking. Finally, smaller subsets of participants completed the MCMI-III and LSRP. Larger samples of offenders would have been preferable for the analyses of these instruments.

Another limitation concerns one of the subscales that emerged from the CTP model, specifically the pattern of correlations between the Grandiosity subscale and the measures used to assess convergent and divergent validity. The Grandiosity subscale was internally consistent, and the construct on which it was based has been identified as relevant to criminal thinking and personality traits such as psychopathy and antisocial personality, yet scores on the subscale were not strongly correlated with the MCMI-III or EQ-i, and only moderately correlated with Factor 1 psychopathy on the LSRP. Grandiosity also had the weakest correlation to the CTP Total Score compared with the other CTP subscales. One possible reason for the observed pattern of correlations may be the type of grandiosity that the subscale measured. There has been some debate regarding subtypes of grandiosity or narcissism. For example, Twenge and Campbell (2009) discussed two types of narcissists, one marked by extraversion and high self-esteem and the other characterized by introversion and low self-esteem, whereas Meloy (2000) discussed an aggressive narcissist whose egocentricity facilitates antisocial behavior. It is possible that the Grandiosity subscale is more closely related to the extraverted or introverted forms of narcissism rather than the more aggressive variant, and thus, is not strongly related to antisocial constructs. Exploring the relationship between the Grandiosity subscale and different forms of narcissism and criminal justice outcomes (e.g., recidivism, institutional misconduct, treatment compliance) will help to clarify whether this scale should be retained as is or revised to reflect the aggressive narcissism described by Meloy.

Several caveats are in order concerning the construct of criminal thinking in general. First, we do not believe that criminal thinking should be the sole focus of offender assessment and rehabilitation. Instead, criminal thinking is one of a host of variables to be considered in the treatment of offenders (Simourd & Olver, 2002). The development and maintenance of criminal behavior is certainly complex and multidetermined. Although thinking patterns may direct behavior, they also emerge from individual experiences, environmental circumstances, and reinforcement histories. From the perspective of the risk-need-responsivity model of offender rehabilitation, criminal thinking is just one of a series of variables to consider in the overall assessment process, including substance use, criminal history, family dynamics, peers, and employment among other factors (Andrews & Bonta, 2010). We are also humbled by the rich literature in social psychology that has consistently demonstrated the important role that situational factors play in shaping antisocial behavior (for a detailed review of this literature see Zimbardo, 2007).

Second, the term criminal thinking may reinforce the idea that offenders are somehow qualitatively different from other people and therefore inherently resistant to change. In contrast to this view, criminal thinking patterns may simply represent a more normative, but maladaptive, cognitive coping style that develops over time with repetition. In the face of life’s challenges and struggles, everyone has the potential to crave excitement, to believe that certain problems cannot be resolved, or to fail to sufficiently consider the impact of one’s actions on the suffering of others. For certain people, such patterns may become more prevalent and automatic, setting the stage for choices, some of which are more likely to lead to criminal behavior. In that sense, we see criminal thinking patterns, such as those profiled on the CTP, as a set of beliefs that affect choices and behaviors, which if unaltered, ultimately influence one’s life trajectory. The next logical question is, “To what extent can each of these attitudes be modified through intervention?”

Criminal Thinking and CBT Interventions

To develop evidence-based interventions for reducing offending behavior, thinking patterns empirically demonstrated to exist in offender groups should establish the foundation of cognitive targets to be modified in forensic CBT programs. Surprisingly, assessments of criminal thinking styles are not generally integrated into real-world treatment and programming for offender-clients (Simourd & Olver, 2002). One reason for this omission may be the divergent origins of criminal thinking assessment instruments and CBT treatment programs; the former originates from the criminology literature and the latter from the cognitive-behavioral psychology treatment literature. It appears that the empirical literature on offender attitudes focuses on thinking patterns as variables that potentially predict program behaviors and recidivism, with less attention dedicated to treatment issues. In contrast, CBT programs from psychology emphasize methods and strategies for altering thinking but tend to highlight thinking patterns central to anxiety and depression rather than criminal behavior. The somewhat disconnected criminology and psychology literatures create a dilemma for practitioners running CBT programs in criminal justice settings. Perhaps the time has come for better integration of thinking patterns, identified by criminal thinking instruments, with real-world forensic CBT programs. Such a union has the potential to provide both relevant thinking targets and effective techniques and strategies to promote change with a traditionally challenging client group.

Given that existing criminal thinking instruments have produced an array of subscales and factors, the question of which specific patterns are central to chronic offending, and most critical to include in intervention programs, has not yet been adequately addressed. The CTP subscale most similar to patterns found on existing criminal thinking instruments is Justifying. One or more subscales concerning the justification or minimization of criminal conduct also appear on the PICTS, CSS-M, TCU CTS, and MCAA. Similarly, intervention programs with offender populations seem to place special emphasis on challenging thinking that minimizes accountability, and having offenders take personal responsibility for their actions seems to be a cognitive focal point (Maruna & Mann, 2006). Despite the popularity of this thinking pattern in criminal thinking assessment and treatment, several recent reviews have suggested that an overemphasis on justification in the offender treatment literature has resulted in the too little attention being paid to other, perhaps more important, cognitive patterns that contribute to offending behavior (Henning & Holdford, 2006; Maruna & Copes, 2005; Maruna & Mann, 2006). The findings in the present study are consistent with these reviews: thinking patterns such as Disregard for Others, Demand for Excitement, and Inability to Cope correlated as highly, or even higher, with psychopathy, aggressive personality disorders, and healthy personality functioning as Justifying. Justification is only a piece of the criminal thinking landscape, and not necessarily the most important thinking pattern. Identifying which specific thinking patterns are most central to offending behaviors seems important for developing a foundation for a general evidence-based treatment protocol. To this end, we see the clinical utility of the CTP as an important avenue for future research.

Future Directions

Future research should focus on further validation of the CTP as well as exploring the utility of the scale to guide cognitive-behavioral interventions. Specifically, a phase of validation should examine the relationship between the CTP and other existing criminal thinking instruments as well as the relationship between the CTP and the recently conceptualized higher order proactive and reactive factors proposed by Walters (2008). Direct comparisons with other instruments such as the PICTS, CSS-M, TCU CTS, and MCAA would provide a more precise picture regarding how CTP subscales converge with those patterns already identified in the offender attitude literature. Additional convergent validation with criminal risk assessment tools such as the Level of Service Inventory–Revised (Andrews & Bonta, 1995) and other psychological instruments commonly used with offenders such as the Minnesota Multiphasic Personality Inventory–2 (Butcher et al., 2001) and the PCL-R (Hare, 2003) should also be investigated. Predictive validation examining the CTP’s relationship with criminal justice variables such as institutional behavior, program/supervision compliance, and recidivism would add to our understanding of the relationship between thinking patterns and behavioral outcomes. Finally, clinically oriented research, exploring effective and efficient methods of altering the thinking patterns, identified on the CTP, and the extent to which reductions in specific patterns are associated with reductions in offending, and improvements in healthy functioning, should be viewed as a practical extension of this work. Treatment research that examines the effectiveness of assessing and pinpointing specific criminal thinking patterns in intervention programs would help advance the development of an empirically supported CBT model for offenders.

Footnotes

Acknowledgements

The authors would like to thank Lisa Angell, Kimberly Bender, Renea Buckwalter, Diane Carlucci, Amanda Hall, Robert Kelly, Lyndsay Ruffalo, Sheila Schwedler, Matthew Tenant, Rachel Tirnady, and the State of Connecticut Court Support Services Division, and the Connecticut Department of Correction for their assistance with this project, as well as the comments of several anonymous reviewers.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research reported in this article was aided by an American Association of University Professors’ University Research Grant and a Central Connecticut State University Dean’s Research Initiative.

References

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders. (4th ed.). Washington, DC: Author.

Andrews

D. A.

Bonta

(1995). The Level of Service Inventory–Revised. Toronto, Ontario, Canada: Multi-Health Systems.

Andrews

D. A.

Bonta

(2010). The psychology of criminal conduct. (4th ed.). New Providence, NJ: Matthew Bard and Company.

Bar-On

(1997). Bar-On Emotional Quotient Inventory. Toronto, Ontario, Canada: Multi-Health Systems.

Beck

A. T.

(1963). Thinking and depression: I. Idiosyncratic content and cognitive distortions. Archives of General Psychiatry, 9, 324-444.

Beck

A. T.

(1967). Depression: Causes and treatment. Philadelphia: University of Pennsylvania Press.

Beck

A. T.

Freeman

Davis

D. D.

Associates. (Eds.). (2004). Cognitive therapy of personality disorders. (2nd ed.). New York, NY: Guilford.

Bentler

P. M.

(2001). EQS 6 structural equations program manual. Encino, CA: Multivariate Software.

Brinkley

C. A.

Schmitt

W. A.

Smith

Newman

J. P.

(2001). Construct validation of a self-report psychopathy scale: Does Levenson’s Self-Report Psychopathy Scale measure the same constructs as Hare’s Psychopathy Checklist–Revised. Personality and Individual Differences, 31, 1021-1038. doi:10.1016/S0191-8869(00)00178-110.1016/S0191-8869(00)00178-1

10.

Bulten

Nijman

van der Staak

(2009). Measuring criminal thinking styles: The construct validity and utility of the PICTS in a Dutch prison sample. Legal and Criminological Psychology, 14, 35-49. doi:10.1348./135532507X25536810.1348./135532507X255368

11.

Butcher

J. N.

Graham

J. R.

Ben-Porath

Y. S.

Tellegen

Dahlstrom

W. G.

Kaemmer

B. K.

(2001). MMPI-2 (Minnesota Multiphasic Personality Inventory–2): Manual for administration, scoring, and interpretation. Minneapolis: University of Minnesota Press.

12.

Butler

Chapman

J. M.

Forman

E. M.

Beck

A. T.

(2006). The empirical status of cognitive-behavioral therapy: A review of meta-analyses. Clinical Psychology Review, 26, 17-31. doi:10.1016/j.cpr.2005.07.00310.1016/j.cpr.2005.07.003

13.

Campbell

French

Gendreau

(2009). The prediction of violence in adult offenders: A meta-analytic comparison of instruments and methods of assessment. Criminal Justice and Behavior, 36, 567-590. doi:10.1177/009385480933361010.1177/0093854809333610

14.

Cleckley

(1964). The mask of sanity. St. Louis, MO: Mosby.

15.

Dembo

Jainhill

Turner

Fong

Farkas

Childs

(2007). Levels of psychopathy and its correlates: A study of incarcerated youths in three states. Behavioral Sciences & the Law, 25, 717-738. doi:10.1002/bsl.78410.1002/bsl.784

16.

Edens

J. F.

(2006). Unresolved controversies concerning psychopathy: Implications for clinical and forensic decision making. Professional Psychology: Research and Practice, 37, 59-65. doi:10.1037/0735-7028.37.1.5910.1037/0735-7028.37.1.59

17.

Edens

J. F.

Poythress

N. G.

Lilienfeld

S. O.

Patrick

(2008). A prospective comparison of two measures of psychopathy in the prediction of institutional misconduct. Behavioral Sciences & the Law, 26, 529-541. doi:10.1002/bsl.82310.1002/bsl.823

18.

Egan

McMurran

Richardson

Blair

(2000). Criminal cognitions and personality: What does the PICTS really measure? Criminal Behaviour and Mental Health, 10, 170-184. doi:10.1002/cbm.35510.1002/cbm.355

19.

Ellis

(1957). Rational psychotherapy and individual psychology. Journal of Individual Psychology, 13, 38-44.

20.

Ellis

(1962). Reason and emotion in psychotherapy. New York, NY: Lyle Stuart.

21.

Gonsalves

V. M.

Scalora

M. J.

Huss

M. T.

(2009). Prediction of recidivism using the Psychopathy Checklist–Revised and the Psychological Inventory of Criminal Thinking Styles within a forensic sample. Criminal Justice and Behavior, 36, 741-756. doi:10.1177/009385480933568810.1177/0093854809335688

22.

Hare

R. D.

(1996). Psychopathy: A clinical construct whose time has come. Criminal Justice and Behavior, 23, 25-54. doi:10.1177/009385489602300100410.1177/0093854896023001004

23.

Hare

R. D.

(2003). The Hare Psychopathy Checklist–Revised manual (2nd ed.). Toronto, Ontario, Canada: Multi-Health Systems.

24.

Harpur

T. J.

Hare

R. D.

Hakstian

(1989). Two-factor conceptualization of psychopathy: Construct validity and assessment implications. Psychological Assessment, 1, 6-17. doi:10.1037/1040-3590.1.1.610.1037/1040-3590.1.1.6

25.

Hemmati

Mills

J. F.

Kroner

D. G.

(2004). The validity of the Bar-On emotional intelligence quotient in an offender population. Personality and Individual Differences, 37, 695-706. doi:10.1016/j.paid.2003.10.00310.1016/j.paid.2003.10.003

26.

Henning

Holdford

(2006). Minimization, denial, and victim blaming by batterers: How much does the truth matter? Criminal Justice and Behavior, 33, 110-130. doi:10.1177/009385480528232210.1177/0093854805282322

27.

Keith

T. Z.

(2006). Multiple regression and beyond. Boston, MA: Pearson.

28.

Knight

Garner

B. R.

Simpson

D. D.

Morey

J. T.

Flynn

P. M.

(2006). An assessment for criminal thinking. Crime & Delinquency, 52, 159-177. doi:10.1177/001112870528174910.1177/0011128705281749

29.

Levenson

M. R.

Kiehl

K. A.

Fitzpatrick

C. M.

(1995). Assesssing psychopathic attributes in a noninstitutionalized population. Journal of Personality and Social Psychology, 68, 151-158. doi:10.1037/0022-3514.68.1.15110.1037/0022-3514.68.1.151

30.

Lilienfeld

S. O.

Fowler

K. A.

(2006). The self-report assessment of psychopathy: Problems, pitfalls, and promises. In Patrick

C. J.

(Ed.), Handbook of psychopathy (pp. 107-132). New York, NY: Guilford.

31.

Mandracchia

J. T.

Morgan

R. D.

Garos

Garland

J. T.

(2007). Inmate thinking patterns: An empirical investigation. Criminal Justice and Behavior, 34, 1029-1043. doi:10.1177/009385480730178810.1177/0093854807301788

32.

Maruna

Copes

(2005). What have we learned in five decades of neutralization research? Crime and Justice: A Review of Research, 32, 221-320.

33.

Maruna

Mann

R. E.

(2006). A fundamental attribution error? Rethinking cognitive distortions. Legal and Criminological Psychology, 11, 155-177. doi:10.1348/135532506X11460810.1348/135532506X114608

34.

Meloy

J. R.

(2000). Violence risk and threat assessment: A practical guide for mental health and criminal justice professionals. San Diego, CA: Specialized Training Services.

35.

Millon

Davis

R. D.

Millon

(1997). Manual for the Millon Clinical Multiaxial Inventory–III (MCMI-III) (2nd ed.). Minneapolis, MN: National Computer Systems.

36.

Mills

J. F.

Kroner

D. G.

Forth

A. E.

(2002). Measures of Criminal Attitudes and Associates (MCAA): Development, factor structure, reliability, and validity. Assessment, 9, 240-253. doi:10.1177/107319110200900300310.1177/1073191102009003003

37.

Mylonas

A. D.

Reckless

W. C.

(1963). Prisoners’ attitudes towards the law and legal institutions. Journal of Criminal Law, Criminology, and Police Science, 54, 479-484.

38.

O’Connor

B. P.

(2000). SPSS and SAS programs for determining the number of components using parallel analysis and Velicer’s MAP test. Behavior Research Methods, Instruments and Computers, 32, 396-402.

39.

Patil

V. H.

McPherson

M. Q.

Friesner

(2010). The use of exploratory factor analysis in public health: A note on parallel analysis as a factor retention criterion. American Journal of Health Promotion, 24, 178-181. doi:10.4278/ajhp.0803313110.4278/ajhp.08033131

40.

Simourd

D. J.

(1997). The Criminal Sentiments Scale–Modified and Pride in Delinquency Scale: Psychometric properties and construct validity of two measures of criminal attitudes. Criminal Justice and Behavior, 24, 52-70. doi:10.1177/009385489702400100410.1177/0093854897024001004

41.

Simourd

D. J.

Hoge

R. D.

(2000). Criminal psychopathy: A risk-and-need perspective. Criminal Justice and Behavior, 27, 256-272. doi:10.1177/009385480002700200710.1177/0093854800027002007

42.

Simourd

D. J.

Olver

M. E.

(2002). The future of criminal attitudes research and practices. Criminal Justice and Behavior, 29, 427-446. doi:10.1177/009385480202900400510.1177/0093854802029004005

43.

Strack

(2008). Essentials of Millon inventories assessment (3rd ed.). Hoboken, NJ: Wiley.

44.

Sutherland

E. H.

Cressey

D. R.

(1978). Criminology (10th ed.). Philadelphia, PA: J. B. Lippincott.

45.

Sykes

G. M.

Matza

(1957). Techniques of neutralization: A theory of delinquency. American Sociological Review, 22, 664-673. doi:10.2307/208919510.2307/2089195

46.

Thompson

(2004). Exploratory and confirmatory factor analysis: Understanding concepts and applications. Washington, DC: American Psychological Association.

47.

Thornton

Blud

(2007). The influence of psychopathic traits on response to treatment. In Herve

Yuille

J. C.

(Eds.), The psychopath: Theory, research, and practice (pp. 505-539). Mahwah, NJ: Lawrence Erlbaum.

48.

Twenge

J. M.

Campbell

W. K.

(2009). The narcissism epidemic: Living in the age of entitlement. New York, NY: Free Press.

49.

Walters

G. D.

(1995). The Psychological Inventory of Criminal Thinking Styles: I. Reliability and preliminary validity. Criminal Justice and Behavior, 22, 307-325. doi:10.1177/009385489502200300810.1177/0093854895022003008

50.

Walters

G. D.

(2005). How many factors are there on the PICTS? Criminal Behaviour and Mental Health, 15, 273-283. doi:10.1002/cbm.2510.1002/cbm.25

51.

Walters

G. D.

(2008). Self-report measures of psychopathy, antisocial personality, and criminal lifestyle: Testing and validating a two-dimensional model. Criminal Justice and Behavior, 35, 1459-1483. doi:10.1177/009385480832092210.1177/0093854808320922

52.

Walters

G. D.

Felix

C. M.

Reinoehl

(2009). Replicability and cross-gender invariance of a two-dimensional model of antisociality in male and female college students. Personality and Individual Differences, 46, 704-708. doi:10.1016/j.paid.2009.01.02810.1016/j.paid.2009.01.028

53.

Walters

G. D.

Geyer

M. D.

(2005). Construct validity of the Psychological Inventory of Criminal Thinking Styles in relationship to the PAI, disciplinary adjustment, and program completion. Journal of Personality Assessment, 84, 252-260. doi:10.1207/s15327752jpa8403_0410.1207/s15327752jpa8403_04

54.

Walters

G. D.

Mandell

(2007). Incremental validity of the Psychological Inventory of Criminal Thinking Styles and Psychopathy Checklist: Screening Version in predicting disciplinary outcome. Law and Human Behavior, 31, 141-157. doi:10.1007/s10979-006-9051-y10.1007/s10979-006-9051-y

55.

Wormith

J. S.

Olver

M. E.

Stevenson

H. E.

Girard

(2007). The long-term prediction of offender recidivism using diagnostic, personality, and risk/need approaches to offender assessment. Psychological Services, 4, 287-305. doi:10.1037/1541-1559.4.4.28710.1037/1541-1559.4.4.287

56.

Wright

E. M.

(2009). The measurement of psychopathy: Dimensional and taxometric approaches. International Journal of Offender Therapy and Comparative Criminology, 53, 464-481. doi:10.1177/0306624X0831941610.1177/0306624X08319416

57.

Yochelson

Samenow

S. E.

(1976). The criminal personality, Volume 1: A profile for change. New York, NY: Jason Aronson.

58.

Zimbardo

P. G.

(2007). The Lucifer effect: Understanding how good people turn evil. New York, NY: Random House.

59.

Zwick

W. R.

Velicer

W. F.

(1986). Comparison of five rules for determining the number of components to retain. Psychological Bulletin, 99, 432-442. doi:10.1037/0033-2909.99.3.43210.1037/0033-2909.99.3.432