Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Apr 5.
Published in final edited form as: J Offender Rehabil. 2022 Mar 15;61(3):135–147. doi: 10.1080/10509674.2022.2045528

A psychometric reevaluation of the TCU criminal thinking scales (CTS)

Thomas B Sease 1, George Joe 1, Jennifer Pankow 1, Wayne E K Lehman 1, Kevin Knight 1
PMCID: PMC8983012  NIHMSID: NIHMS1792050  PMID: 35386231

Abstract

In the United States, approximately 9 million people cycle in and out of jail and more than 600,000 people are released from prison each year. Unfortunately, the reentry process includes several barriers people must overcome (e.g., criminal thinking) to achieve adequate psychosocial functioning. As such, valid and reliable assessments that allow correctional staff to monitor clients’ progress in treatment and test program effectiveness are paramount to reducing this major public safety concern. The TCU Criminal Thinking Scales (CTS) are a widely used assessment of criminal thinking in correctional settings. This study reevaluated the psychometric properties of the TCU CTS using Item Response Theory. Results showed the TCU CTS had good internal reliability and each scale loaded onto one factor. Item level analysis revealed most items adequately fit the model, generally measuring moderate levels of criminal thinking. Furthermore, several TCU CTS scales were negatively correlated with motivation for treatment and psychosocial functioning.

Keywords: assessment, criminal thinking, evaluation, criminal justice system, measurement

Introduction

In the United States, approximately 9 million people cycle in and out of jail and more than 600,000 people are released from prison each year (U.S. Department of Health & Human Services, 2021). Unfortunately, the reentry process includes contextual barriers justice-involved people must overcome, such as finding employment, housing, and transportation (see Bushway et al., 2007 for a full review). Exacerbating these concerns are maladaptive cognitive patterns or negative attitudes that interfere with psychosocial functioning. Together, difficulties associated with the reentry process place this population at an elevated risk for subsequent involvement in the criminal justice system. A longitudinal study conducted by the U.S. Department of Justice found that more than three-fourths (83%) of people released from state prisons recidivated within nine years of release (Alper et al., 2018). To address this public safety challenge, considerable effort has been devoted to providing justice-involved people with services that reduce their likelihood of criminal activity post-release.

The Risk-Need-Responsivity (RNR; Andrews & Bonta, 2010) model of risk assessment and treatment needs posits that services with the greatest potential to reduce recidivism are those that (1) match the intensity of care with the level of individual need, (2) target the client’s unique needs as the mechanism of change, and (3) provide support that maximizes a person’s likelihood of benefiting from the intervention. A fundamental component of this approach is mitigating criminogenic factors that contribute to a person’s risk for recidivism. This is performed under the assumption that attenuating maladaptive social, psychological, and personal patterns of behavior will reduce a person’s probability of reoffending. Noteworthy criminogenic factors amenable through intervention are antisocial personality patterns (e.g., impulsivity, aggressiveness, irritability) and pro-criminal attitudes (Bonta & Andrews, 2007). Antisocial personality patterns have been connected to criminal behavior in prospective research, and pro-criminal attitudes (e.g., Personal Irresponsibility, Justification) have been correlated with substance use among youth in the juvenile justice system (Dembo et al., 2007; Fridell et al., 2008). Similarly, decreases in criminal thinking have been associated with fewer disciplinary infractions and less prison misconduct in justice-involved populations (Folk et al., 2016; Walters, 2017).

Meta-analytic strategies have consistently demonstrated a connection between criminal thinking and criminal behavior (Walters, 2002, 2012). For example, criminal thinking predicted recidivism even while controlling for criminal history (Walters, 2012). Relatedly, criminal thinking was positively correlated with justice-involved individuals’ history of lifetime arrests and negatively correlated with recovery attitudes (Bartholomew et al., 2018). The literature has identified several status factors linked to criminal thinking, such as more education, a longer sentence length, and more time served (Mandracchia & Morgan, 2010). Justice-involved males tend to report higher levels of criminal thinking when compared to their female counterparts (Taxman et al., 2011). Studies examining the proximal outcomes of criminal thinking, in combination with those investigating personal factors related to criminal thinking, provide critically important information regarding intervention opportunities that have the potential to improve longer-term public safety and health outcomes.

An easy to administer and cost-effective measure of criminal thinking in justice-involved populations is the 36-item Criminal Thinking Scales (CTS; Knight et al., 2006). The TCU CTS was initially developed for the evaluation of cognitive-oriented substance use treatment programs. The TCU CTS includes the following six scales: Entitlement (EN), Justification (JU), Criminal Rationalization (CN) Personal Irresponsibility (PI), Power Orientation (PO) and Coldheartedness (CH). The administration takes less than 15 minutes, and the scoring of the scales is achieved by computing the average of items responses within each scale, minimizing the challenges often associated with the administration and scoring of screening instruments used with justice-involved populations. In this way, the TCU CTS offers correctional staff an easy, cost-effective, and easily interpretable tool that can be used to monitor client progress in treatment.

The psychometrics of the TCU CTS have been established using Classical Test Theory (e.g., Knight et al., 2006). In the original paper, all scales of the TCU CTS were found to have acceptable internal reliability scores (α = .68–.81) and scale structures. Subsequent investigations have supported the validity of TCU CTS, with criminal thinking being correlated with less engagement in substance use treatment, worse client functioning, and decreased adherence to treatment (Best et al., 2009; Dembo et al., 2007; Mitchell et al., 2013; Sana & Batool, 2017). Personal Irresponsibility and Justification were positively associated with crack cocaine use severity, and Personal Irresponsibility was associated with more cocaine use dependence and weekly drug use (Packer et al., 2009). Criminal thinking is not only associated with increased substance use severity but also decreased psychological well-being (Best et al., 2009; Butt et al., 2019).

Although the TCU CTS scale is a valid and reliable assessment of criminal thinking, advancements in scale development make it possible for researchers to assess the psychometric properties of a scale using item-level analyses (i.e., Item Response Theory; IRT); these contrast measurement-level analyses used by Classical Test Theory. This process allows for the assessment of each item’s validity within a construct and the removal of ill-fitting items without diminishing the overall validity of the assessment. These procedures provide researchers additional information about the scale, which can be used to refine assessments with greater detail. To this end, the present study focuses on a psychometric reevaluation of the TCU CTS using IRT.

Current study

This study examined the psychometric properties of each TCU CTS scale using IRT procedures. This was achieved using exploratory and confirmatory factor analysis, followed by an examination of internal reliability scores, item fit, and difficulty scores. Furthermore, this study evaluated the validity of the TCU CTS by examining its association with motivation for treatment (e.g., desire for help, problem recognition, treatment readiness) and psychosocial functioning (e.g., depression, anxiety, self-esteem). The authors expected that criminal thinking scores to be negatively associated with treatment motivation and self-esteem while being positively associated with measures of depression and anxiety.

Method

Participants and procedure

This study was a data reanalysis using secondary data collected from eight prison-based substance use disorder treatment programs that were part of the TCU Disease Risk Reduction (DRR) Project (see Lehman et al., 2015). Five male-only and three female-only units from two different states were included in the study. Treatment programs were modified therapeutic community programs wherein clients were required to receive a minimum of 20 h of programing each week. All new intakes completed a battery of forms as part of routine clinical practice which included the TCU Global Risk assessment Adult version (TCU A-RSKForm), the TCU Drug Screen II, TCU Client Evaluation of Self in Treatment (CEST) Forms, and the TCU Criminal Thinking Scales (CTS). All data were de-identified and analyzes were performed on responses provided at time-point one, making this study cross-sectional. All procedures associated with this study were approved by the authors’ Institutional Review Board.

The resultant sample consisted of 8,351 justice-involved males (67.1%) and females (32.9%) ranging in age from 17 to 71 (M = 34.69, SD = 9.78). Half of the sample reported their race as White (50.7%), followed by Black/African American (25.6%) or Other (23.7%). The majority of respondents identified as non-Hispanic (81.2%) and only a quarter (24.1%) were married. According to the TCU Drug Screen II (Knight et al., 2002), about two-thirds (63.1%) of the sample met the criteria for a severe substance use problem.

Measures

Criminal thinking

The 36-item TCU CTS was used to assess criminal thinking in the following six dimensions: Personal Irresponsibility, Coldheartedness, Criminal Rationalization, Power Orientation, Entitlement, and Justification. Using a 5-point Likert scale (1 = Strongly Disagree, 5 = Strongly Agree), participants rated items such as, “When people tell you what you do, you become aggressive,” “You are not to blame for everything you have done,” and “You feel that you are above the law” for Power orientation, Personal Irresponsibly, and Entitlement, respectively. In previous work, all scales on the TCU CTS have demonstrated strong psychometric properties (Knight et al., 2006).

Motivation for change

The TCU MOTForm (Simpson & Joe, 1993) assessed desire for help, problem recognition, and treatment readiness. Using a 5-point Likert scale (1 = Strongly Disagree, 5 = Strongly Agree), participants were asked to report on items about their motivation for substance use treatment. Sample items include, “Your drug use is a problem for you,” “You need help dealing with your drug use,” and “You want to be in drug treatment” for problem recognition, desire for help, and treatment readiness, respectively. Extant literature indicates these scales are valid and reliable in justice-involved populations (Pankow et al., 2012).

Psychological functioning

To measure psychological functioning, the self-esteem, depression, and anxiety scales on the TCU PSYForm were used. Respondents are asked to rate how much they agreed or disagreed with each item using a 5-point Likert scale (1 = Strongly Disagree, 5 = Strongly Agree). Example items include, “You have a lot to be proud of,” “You feel hopeless about the future,” and “You feel tense or keyed-up” for self-esteem, depression, and anxiety, respectively. The TCU PSYForm has demonstrated strong psychometric properties in justice-involved populations (Pankow et al., 2012).

Analytic plan

The reanalysis of the TCU CTS was performed using the concepts from IRT (see De Ayala, 2013 for a comprehensive review of IRT). First, descriptive statistics (M, SD) were computed for each TCU CTS scale (see Table 1). Next, each CTS scale was examined to ensure the assumptions of IRT were met (e.g., unidimensionality, monotonicity). The assumption of local independence was not tested because it was assumed that the correlation of the residuals between any two items would be uncorrelated after the trait being measured by the items was partitioned out. Finally, one random subsample of 400 participants was obtained from the full dataset to test for construct validity. All analyses listed herein were completed using Winsteps Version 4.4.6, SAS Version 9.4, and SPSS Version 25.0.

Table 1.

Descriptive statistics for criminal thinking scores.

Mean SD 33rd percentile 67th percentile
Cold Heartedness 2.16 0.60 2.00 2.40
Personal Irresponsibility 2.03 0.61 1.83 2.17
Power Orientation 2.32 0.71 2.00 2.57
Justification 2.01 0.63 1.83 2.17
Criminal Rationalization 2.84 0.77 2.50 3.17
Entitlement 1.73 0.56 1.33 2.00
Open in a new tab

Note. Scores for each scale range from 1 (Strongly Disagree) to 5 (Strongly Agree). SD: standard deviation.

Item response theory (IRT)

Exploratory and confirmatory factor analysis was used to test the latent structure of each scale and evaluate the assumption of unidimensionality. Factors with an eigen value greater than one were retained (Kaiser, 1960) and internal reliability scores were computed. Confirmatory factor analysis was used to verify the structure of each CTS scale and test each model using indicators of model fit (e.g., χ2, GFI, SRMR, RMSEA). Then, item fit was evaluated using the point-measure correlations, mean-squared fit statistics (MNSQ), and difficulty scores provided in Winsteps. Items with a point-measure correlation greater than 0.40 were considered to have an acceptable fit (Kean et al., 2018). Likewise, items with a MNSQ infit and outfit score between 0.60 and 1.4 considered well-fitting (Kean et al., 2018). The former provides an estimate of model fit that is sensitive to unusual scores toward the middle the sampling distribution whereas the latter is sensitive to unusual scores at either tail of the distribution. Finally, items were evaluated using the difficulty score for each item within their respective scale. Difficulty scores represent the average ability levels (i.e., degree of criminal thinking) captured by a particular item. Higher difficulty scores suggest that people endorsing these items are more likely to have higher “criminal thinking” whereas people only endorsing items with low difficulty are more likely to have lower “criminal thinking.”

Validity

To evaluate the construct validity of the TCU CTS, correlation analysis determined the associations among the TCU CTS scales, motivation for treatment, and psychosocial functioning.

Results

Dimensionality

Using the full sample (N = 8,351), exploratory factor analysis with a varimax rotation tested the latent structure of each scale on the TCU CTS. Results showed each of the scales loaded best onto one factor, all achieving eigenvalues greater than 2 and explaining a large portion of the observed variance in their respective latent construct (≥39.49%; see Table 2 for factor loading). All scales showed acceptable internal reliability, with Cronbach alphas of 0.83, 0.76, 0.82, 0.64, 0.75, and 0.68 for EN, JU, PO, CH, CN, and PI, respectively.

Table 2.

Model of fit indices for each CTS subscale.

Scale χ 2 SRMR RMSEA GFI
Cold Heartedness <.001 0.02 0.05 0.99
Personal Irresponsibility <.001 0.02 0.03 0.99
Power Orientation <.001 0.03 0.07 0.98
Justification <.001 0.03 0.08 0.98
Criminal Rationalization <.001 0.01 0.02 0.99
Entitlement <.001 0.03 0.08 0.97
Open in a new tab

Note. SRMR: Standardized Root Mean Square Residual; RMSEA: Root Mean Square Error of Approximation; GFI: Goodness of Fit Index.

Confirmatory factor analysis tested the model fit of each individual TCU CTS scale when being forced onto a single factor. Results showed that the minimum fit χ2 test was significant (ps < .001), which was expected considering the large sample size. Follow-up tests using other estimation methods of model fit (e.g., GFI, SRMR, RMSEA) showed that each TCU CTS scale achieved acceptable model fit when examined as a single factor (see Table 2).

Item fit

Results from the point-measure correlations showed that all items were positively correlated with each other, exceeding the 0.40 criteria for determining acceptable item fit. In addition, MNSQ infit and outfit statistics showed that all items except Item 15 adequately fit the model for their respective scale (see Table 3). The MNSQ outfit for Item 15 was 1.59. When examining difficulty values, the item with the highest difficulty score was Item 3 (i.e., “The real reason you are locked-up is because of your race”) and the item with the lowest difficulty score was Item 15 (i.e., “You like to be in control”). In general, results showed that the TCU CTS scales are primarily composed of items with moderate difficulty.

Table 3.

Item fit statistics.

Item # Infit MNSQ Infit ZSTD Outfit MNSQ Outfit ZSTD PTMZ R corr. Difficulty scores
CH
 1 1.08 4.35 1.08 4.05 0.62 −0.04
 6 0.87 −7.97 0.87 −7.49 0.65 −0.25
 12 1.07 4.38 1.17 9.61 0.62 −0.69
 17 1.21 9.48 1.16 7.70 0.54 0.74
 27 0.83 −9.40 0.81 −9.90 0.61 0.23
PI
 2 1.11 6.68 1.22 9.90 0.59 −0.59
 3 1.08 3.91 0.93 −3.05 0.48 1.28
 21 1.06 3.39 1.19 9.87 0.55 −0.26
 29 0.99 −0.74 1.08 4.35 0.60 −0.43
 31 0.79 −9.90 0.79 −9.90 0.61 0.09
 36 1.04 1.92 1.05 2.56 0.59 −0.09
PO
 4 0.94 −3.63 0.96 −2.22 0.66 0.29
 10 0.85 −8.98 .84 −9.35 0.67 0.44
 13 1.09 5.42 1.17 9.14 0.62 −0.06
 14 0.89 −7.02 0.86 −8.06 0.71 0.16
 15 1.37 9.90 1.59 9.90 0.61 −1.23
 20 0.82 −9.90 0.82 −9.90 0.69 0.29
 28 1.08 4.46 1.08 4.49 0.66 0.10
JU
 7 1.03 1.74 1.17 9.51 0.62 −0.50
 11 1.03 1.67 1.17 9.53 0.65 −0.57
 16 1.05 2.75 1.06 2.97 0.63 0.02
 25 1.13 5.74 1.03 1.55 0.57 0.76
 26 0.94 −2.92 0.93 −3.55 0.66 0.10
 35 0.93 −3.79 0.89 −5.71 0.64 0.19
CN
 5 1.12 7.87 1.12 7.54 0.63 0.16
 8 0.84 −9.90 0.85 −9.90 0.70 −0.54
 18 1.40 9.90 1.44 9.90 0.54 0.17
 19 0.77 −9.90 0.82 −9.90 0.67 −0.34
 30 1.02 1.50 0.98 −1.29 0.63 0.57
 34 0.86 −9.90 0.90 −6.96 0.67 −0.02
EN
 9 0.99 −0.26 1.09 3.74 0.69 −0.28
 22 1.07 2.79 0.98 −1.02 0.69 0.55
 23 1.05 1.83 0.92 −3.17 0.71 0.37
 24 1.13 5.32 1.16 6.42 0.70 −0.27
 32 1.07 3.18 1.27 9.90 0.68 −0.65
 33 0.79 −8.99 0.70 −9.90 0.74 0.28
Open in a new tab

Note. MNSQ: mean-square; ZSTD: standardized weighted (infit) and unweighted (outfit) mean-squared fit statistics; PTMZ: point-measure correlation.

Validity

Using a random sample of 400 respondents, correlation analysis tested the validity of the TCU CTS. As shown in Table 4, most TCU CTS scales were associated in a positive direction. The exception to this trend was the CH scale, which was not associated with the PO, JU, or CN scales.

Table 4.

Bivariate correlations.

CH PI PO JU CN EN AX SE DP DH TR
CH
PI .17*
PO .02 .49**
JU .05 .64** .69**
CN .01 .51** .44** .42**
EN .21** .70** .59** .73** .39**
AX −.12* .28** .43** .37** 0.31** .28**
SE .04 −.18** −.32** −.29** −.18** −.16** −49**
DP −.03 .35** .45** .39** .32** .29** .71** −.59**
DH −.23** −.26** −.02 .01 −.10* −.16** .20** −.28** .27**
TR −.21** −.31** −.11* −.04 −.22** −.17** .04 −.13** −.02 .72**
PR −.17* −.17* .05 .12* −.05 −.04 .25** −.32** .22** .84** .68**
Open in a new tab

Note. CH: Cold Heartedness; PI: Personal Irresponsibility; PO: Power Orientation; JU: Justification; CN: Criminal Rationalization; EN: Entitlement; AX: Anxiety; SE: Self-Esteem; DP: Depression; DH: Desire for Help; TR: Treatment Readiness; PR: Problem Recognition.

**

p < .01,

*

p < .05.

Most TCU CTS scales were negatively correlated with desire for help, problem recognition, and treatment readiness. That is, higher TCU CTS scores were related to lower motivation for treatment. All scales except CH were negatively associated with self-esteem and multiple scales were positively associated with anxiety and depression. These results indicate that CTS scores were related to lower psychological well-being in this sample of justice involved individuals.

Discussion

This study reevaluated the psychometric properties of the TCU CTS. Exploratory factor analysis supported a one-factor solution for each TCU CTS scale and confirmatory factor analysis suggested the proposed one-factor model for each scale appropriately fit the data. Reliability coefficients ranged from 0.64 to 0.83, indicating these scales can be used as reliable measures of criminal thinking. Next, items were evaluated within their respective scale using indicators of item fit. Results showed most items achieved an acceptable item fit score, as indicated by point-measure correlations, MNSQ infit scores, and MNSQ outfit scores. In evaluating difficulty scores, TCU CTS scales most commonly contained a high number of items with moderate difficulty. This trend across scales implies that the items on the TCU CTS are most adept in capturing information among people with moderate levels of criminal thinking.

This study also revealed that measurements of criminal thinking were negatively associated with desire for help, problem recognition, and treatment readiness—key indicators of motivation for treatment. Motivation for changing substance use behavior has been linked to improved treatment satisfaction and client retention (Joe et al., 1998; Sia et al., 2000). As such, criminal thinking could be a barrier for clients entering substance use treatment, potentially predicting clients’ responsiveness to treatment. The TCU CTS scales were also correlated with variables measuring psychosocial functioning (i.e., depression, anxiety, self-esteem). These findings converge with existing literature reporting a negative association between criminal thinking and psychological well-being (Best et al., 2009; Butt et al., 2019). Together, criminogenic cognitions are associated with decreased motivation for treatment and psychosocial functioning. These findings support the need for evidence-based services that ameliorate criminal thinking patterns.

Findings provide further psychometric support for the use of the TCU CTS as a valid and reliable assessment of criminogenic cognitions. In accordance with the TCU treatment model (Simpson, 2000), criminal thinking measures would ideally be used in two ways. First, for people entering substance use treatment, assessments of criminal thinking can be utilized in conjunction with other validated screeners to examine the client’s problem severity. As demonstrated here, people with high levels of criminal thinking may concurrently experience low motivation for change and psychosocial functioning. This information could be used to make treatment decisions intended to match the level of care with the amount of need in the individual. Second, the TCU CTS scales should be used to monitor client progress in treatment as a pre-post assessment survey. Applications of the TCU treatment model have demonstrated early recovery is associated with treatment adherence, which in turn improves treatment outcomes (Simpson & Joe, 2004). Early recovery in treatment could be measured using the TCU CTS wherein positive changes in criminal thinking would be expected to be linked to improved client outcomes.

While these findings support the utility of the TCU CTS, ensuing investigations could focus on improving these measures. Most items on the TCU CTS had moderate difficulty scores. Future iterations of the scale might consider including items intended capture low and high ability levels. This could include adding items that have a high probability of being selected by most respondents as well as items with a low probability of being highly endorsed. Future studies may also consider exploring the feasibility of creating a TCU CTS that has the capacity to be scored as a composite scale. While the individual scales demonstrated strong psychometric properties in this study, a composite scale could be of practical importance; providing correctional staff and treatment providers a single score for risk classification. Finally, the TCU CTS would benefit from a more comprehensive investigation of its validity. This study showed the TCU CTS scales were more highly correlated with psychosocial functioning variables (e.g., self-esteem, depression, anxiety) than motivation for change. As a result, the TCU CTS could be measuring aspects of criminal thinking that are more relevant to a person’s overall well-being than potential responsiveness to interventions for substance use. Succeeding investigations are needed to explore this possibility.

There are several limitations of the present study. This study focused on reevaluating the psychometrics of the TCU CTS using a sample of justice-involved individuals incarcerated at eight different correctional facilities. These data do not provide information about the correlates of criminal thinking for people involved in the criminal justice system in other ways (e.g., probation, parole). Such information is warranted considering one investigation found criminal thinking to be correlated with recidivism for people in jail, but not for people on probation (Caudy et al., 2015). Data used for this study was cross-sectional, measured at prison intake. Thus, associations between variables do not imply causal relationships. Lastly, this study only investigated the associations between the TCU CTS scales, motivation for change, and psychosocial functioning. This approach fails to account for potential mediators or moderators that may be influencing these relationships. Studies should continue to investigate mediators and moderators of criminal thinking and its distal outcomes to help elucidate how and when criminal thinking is most predictive of negative psychosocial outcomes.

In summation, a primary aim of the criminal justice system is to provide services that reduce clients’ risk for future criminal activity, involvement in the criminal justice system, and recidivism. Equally, these services aim to provide people involved in the justice system the skills needed to succeed as a member of society following reentry. Criminal thinking is a dynamic cognitive process malleable through an intervention that, if accurately measured, could serve as a crucial component of an individual’s treatment plan.

References

  1. Alper M, Durose MR, & Markman J (2018). 2018 update on prisoner recidivism: A 9-year follow-up period (2005–2014). US Department of Justice, Office of Justice Programs, Bureau of Justice Statistics. https://babsim.com/wp-content/uploads/2019/04/2018-update-on-prisoner-recidivism.pdf [Google Scholar]
  2. Andrews DA, & Bonta J (2010). Rehabilitating criminal justice policy and practice. Psychology, Public Policy, and Law, 16(1), 39–55. 10.1037/a0018362 [DOI] [Google Scholar]
  3. Bartholomew NR, Morgan RD, Mitchell SM, & Van Horn SA (2018). Criminal thinking, psychiatric symptoms, and recovery attitudes among community mental health patients: An examination of program placement. Criminal Justice and Behavior, 45(2), 195–213. 10.1177/0093854817734007 [DOI] [Google Scholar]
  4. Best D, Day E, Campbell A, Flynn PM, & Simpson DD (2009). Relationship between drug treatment engagement and criminal thinking style among drug-using offenders. European Addiction Research, 15(2), 71–77. 10.1159/000189785 [DOI] [PubMed] [Google Scholar]
  5. Bonta J, & Andrews DA (2007). Risk-need-responsivity model for offender assessment and rehabilitation. Rehabilitation, 6(1), 1–22. https://www.securitepublique.gc.ca/cnt/rsrcs/pblctns/rsk-nd-rspnsvty/rsk-nd-rspnsvty-eng.pdf [Google Scholar]
  6. Bushway SD, Stoll MA, & Weiman D (Eds.). (2007). Barriers to reentry? The labor market for released prisoners in post-industrial America. Russell Sage Foundation. https://books.google.com/books?hl=en&lr=&id=YOeFAwAAQBAJ&oi=fnd&pg=PR7&dq=Barriers+to+Reentry%3F+The+Labor+Market+for+Released+Prisoners+in+Post-Industrial+America&ots=SsnPF_jFTq&sig=4fVz_6DZ9L0OraS0ldxDf_y0mO0#v=onepage&q=Barriers%20to%20Reentry%3F%20The%20Labor%20Market%20for%20Released%20Prisoners%20in%20Post-Industrial%20America&f=false [Google Scholar]
  7. Butt A, Abdi SK, Hamid A, Dogar FA, & Fatima J (2019). Criminal thinking, moral disengagement and psychological wellbeing in prisoners. Journal of Fatima Jinnah Medical University, 13(2), 51–54. https://www.jfjmu.com/index.php/ojs/article/view/590 [Google Scholar]
  8. Caudy MS, Folk JB, Stuewig JB, Wooditch A, Martinez A, Maass S, Tangney JP, & Taxman FS (2015). Does substance misuse moderate the relationship between criminal thinking and recidivism? Journal of Criminal Justice, 43(1), 12–19. 10.1016/j.jcrimjus.2014.11.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. De Ayala RJ (2013). The theory and practice of item response theory. Guilford Publications. [Google Scholar]
  10. Dembo R, Turner CW, & Jainchill N (2007). An assessment of criminal thinking among incarcerated youths in three states. Criminal Justice and Behavior, 34(9), 1157–1167. 10.1177/0093854807304348 [DOI] [Google Scholar]
  11. Folk JB, Disabato DJ, Daylor JM, Tangney JP, Barboza S, Wilson JS, Bonieskie L, & Holwager J (2016). Effectiveness of a self-administered intervention for criminal thinking: Taking a chance on change. Psychological Services, 13(3), 272–282. 10.1037/ser0000079 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fridell M, Hesse M, Jaeger MM, & Kühlhorn E (2008). Antisocial personality disorder as a predictor of criminal behavior in a longitudinal study of a cohort of abusers of several classes of drugs: Relation to type of substance and type of crime. Addictive Behaviors, 33(6), 799–811. 10.1016/j.addbeh.2008.01.001 [DOI] [PubMed] [Google Scholar]
  13. Joe GW, Simpson DD, & Broome KM (1998). Effects of readiness for drug abuse treatment on client retention and assessment of process. Addiction, 93(8), 1177–1190. 10.1080/09652149835008 [DOI] [PubMed] [Google Scholar]
  14. Kaiser HF (1960). The application of electronic computers to factor analysis. Educational and Psychological Measurement, 20(1), 141–151. 10.1177/001316446002000116 [DOI] [Google Scholar]
  15. Kean J, Bisson EF, Brodke DS, Biber J, & Gross PH (2018). An introduction to item response theory and Rasch analysis: Application using the eating assessment tool (EAT-10). Brain Impairment, 19(1), 91–102. 10.1017/BrImp.2017.13 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Knight K, Garner BR, Simpson DD, Morey JT, & Flynn PM (2006). An assessment for criminal thinking. Crime & Delinquency, 52(1), 159–177. 10.1177/0011128705281749 [DOI] [Google Scholar]
  17. Knight K, Simpson DD, & Morey JT (2002). An evaluation of the TCU drug screen. National Institute of Justice, Office of Justice Programs. U.S. Department of Justice. https://www.ojp.gov/pdffiles1/nij/grants/196682.pdf [Google Scholar]
  18. Lehman WE, Rowan GA, Greener JM, Joe GW, Yang Y, & Knight K (2015). Evaluation of WaySafe: A disease-risk reduction curriculum for substance-abusing offenders. Journal of Substance Abuse Treatment, 58, 25–32. 10.1016/j.jsat.2015.05.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Mandracchia JT, & Morgan RD (2010). The relationship between status variables and criminal thinking in an offender population. Psychological Services, 7(1), 27–33. 10.1037/a0016194 [DOI] [Google Scholar]
  20. Mitchell D, Tafrate RC, Hogan T, & Olver ME (2013). An exploration of the association between criminal thinking and community program attrition. Journal of Criminal Justice, 41(2), 81–89. 10.1016/j.jcrimjus.2012.09.003 [DOI] [Google Scholar]
  21. Packer G, Best D, Day E, & Wood K (2009). Criminal thinking and self-control among drug users in court mandated treatment. Criminology & Criminal Justice, 9(1), 93–110. 10.1177/1748895808099182 [DOI] [Google Scholar]
  22. Pankow J, Simpson DD, Joe GW, Rowan-Szal GA, Knight K, & Meason P (2012). Examining concurrent validity and predictive utility for the Addiction Severity Index and Texas Christian University (TCU) short forms. Journal of Offender Rehabilitation, 51(1–2), 78–95. 10.1080/10509674.2012.633021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Sana F, & Batool I (2017). Development and validation of an indigenous criminal thinking scale. Pakistan Journal of Psychological Research, 32(1), 117–139. http://www.pjprnip.edu.pk/index.php/pjpr/article/view/605/523 [Google Scholar]
  24. Sia TL, Dansereau DF, & Czuchry ML (2000). Treatment readiness training and probationers’ evaluation of substance abuse treatment in a criminal justice setting. Journal of Substance Abuse Treatment, 19(4), 459–467. 10.1016/S0740-5472(00)00139-2 [DOI] [PubMed] [Google Scholar]
  25. Simpson DD, (2000). Research summary: Focus on treatment process and outcomes. TCU Model of treatment process and outcomes Institute of Behavioral Research, Texas Christian University. https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.80.3414&rep=rep1&type=pdf [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Simpson DD, & Joe GW (1993). Motivation as a predictor of early dropout from drug abuse treatment. Psychotherapy: Theory, Research, Practice, Training, 30(2), 357–368. 10.1037/0033-3204.30.2.357 [DOI] [Google Scholar]
  27. Simpson DD, & Joe GW (2004). A longitudinal evaluation of treatment engagement and recovery stages. Journal of Substance Abuse Treatment, 27(2), 89–97. 10.1016/j.jsat.2004.03.001 [DOI] [PubMed] [Google Scholar]
  28. Taxman FS, Rhodes AG, & Dumenci L (2011). Construct and predictive validity of criminal thinking scales. Criminal Justice and Behavior, 38(2), 174–187. 10.1177/0093854810389550 [DOI] [Google Scholar]
  29. U.S. Department of Health & Human Services. (2021). Incarceration & reentry. Office of the Assistant Secretary for Planning and Evaluation. https://aspe.hhs.gov/incarceration-reentry [Google Scholar]
  30. Walters GD (2002). The Psychological Inventory of Criminal Thinking Styles (PICTS): A review and meta-analysis. Assessment, 9(3), 278–291. 10.1177/1073191102009003007 [DOI] [PubMed] [Google Scholar]
  31. Walters GD (2012). Criminal thinking and recidivism: Meta-analytic evidence on the predictive and incremental validity of the Psychological Inventory of Criminal Thinking Styles (PICTS). Aggression and Violent Behavior, 17(3), 272–278. 10.1016/j.avb.2012.02.010 [DOI] [Google Scholar]
  32. Walters GD (2017). Effect of a brief cognitive behavioural intervention on criminal thinking and prison misconduct in male inmates: Variable-oriented and person-oriented analyses. Criminal Behaviour and Mental Health, 27(5), 457–469. 10.1002/cbm.2028 [DOI] [PubMed] [Google Scholar]

RESOURCES