Black–White and Male–Female Differences in Criminal Thinking: Examining Instrumental and Expressive Motives for Crime in Federal Supervisees

Abstract

Using 2,000 federal supervisees comprised of 500 White males, 500 Black males, 500 White females, and 500 Black females, this study evaluated whether race and sex are differentially associated with proactive and reactive criminal thinking. It was predicted that proactive criminal thinking would be higher in Black than White supervisees and that reactive criminal thinking should be higher in female than male supervisees. Results revealed that instrumental motives for crime, as represented by proactive criminal thinking, were more prevalent in Black male offenders, and expressive motives for crime, as represented by reactive criminal thinking, were more prevalent in White female offenders.

Keywords

race gender proactive criminal thinking reactive criminal thinking

Criminal cognition is mentioned sporadically in the criminological literature, but usually only within the context of mainstream criminological theories like Cornish and Clarke’s (1986) rational choice theory or Sykes and Matza’s (1957) techniques of neutralization. This, despite the fact, Andrews and Bonta (2010) considered criminal cognition to be one of the “Big Four” predictors of recidivism. One factor that may be standing in the way of general acceptance of criminal thinking as a core criminological construct is the absence of a generally accepted or compelling theory of criminal cognition (Bernard, Snipes, & Gerould, 2010). Early attempts to create a cognitive theory of crime, such as Yochelson and Samenow’s (1976) criminal personality and Kohlberg’s (1981) moral disengagement perspective, were limited by conceptual and methodological problems. Walters’s (1990) criminal lifestyle model also has a strong cognitive emphasis and with it an instrument, the Psychological Inventory of Criminal Thinking Styles (PICTS; Walters, 1995), that can be used to assess various facets of criminal thinking. This model also assesses cognitive-motivational factors that have been found to be effective in differentiating and testing theories of crime (Agnew, 1995).

Criminal thinking can be viewed as a set of intervening variables capable of connecting key independent variables from core criminological theories to various categories of criminal conduct. The PICTS, in fact, produces scores on two general dimensions of criminal thinking that may be of cardinal significance in mediating various relationships between criminological variables. These are referred to by Walters (2012) as proactive and reactive criminal thinking. Whereas proactive criminal thinking is planned, calculated, and reflective of a fearless temperament, reactive criminal thinking is impulsive, emotional, and reflective of a disinhibited temperament (Walters, 2015). Although different, the two dimensions overlap extensively in offender populations, with intercorrelations ranging from .50 to .70 (Walters, 2002, 2007, 2012, in press). Despite the overlap, the two dimensions frequently exhibit countervailing relationships with outside criteria. As a case in point, proactive criminal thinking has been found to correlate with positive outcome expectancies for crime and a history of instrumental crimes like robbery and burglary, whereas reactive criminal thinking has been found to correlate with hostile attribution biases and a history of expressive crimes like assault and domestic violence (Walters, 2007; Walters, Frederick, & Schlauch, 2007). More recently, Walters (2016) found that proactive criminal thinking mediated the peer influence effect (delinquent peer associations leading to delinquency), whereas reactive criminal thinking mediated the peer selection effect (delinquency leading to delinquent peer associations).

Demographic characteristics such as race and gender have not been widely studied with respect to criminal thinking, and the research that has been conducted thus far has revealed minimal relationship between these two demographic measures and scores on the PICTS (see Walters, 2014). As independent variables with relevance to criminal offending, however, race and gender require additional scrutiny in studies on criminal thinking. After all, criminal thinking can be considered a cognitive-motivational factor with the capacity to mediate the effects of independent variables like race and sex on crime along the lines described by Agnew (1995). Black–White racial status, for example, has been found to interact with cognition in relationship to crime. In an early study on this issue, Harris and Lewis (1974) discovered that Black male inmates in a New Jersey state correctional facility viewed crime as enhancing their self-worth, whereas White male inmates from the same institution viewed crime as detracting from their self-worth. Three decades later, Owens-Sabir (2005) observed a positive relationship between self-esteem and self-reported delinquency in Black but not White youth. In a study restricted to female delinquents, however, Holsinger and Holsinger (2005) failed to uncover evidence of a differential relationship between self-esteem and delinquency in Black versus White girls. Most recently, Walters (2011) determined that positive outcome expectancies for crime were significantly stronger in Black male inmates compared with White male inmates after age, criminal offense, general criminal thinking, and negative outcome expectancies for crime had been controlled. These results suggest that Black respondents, at least Black male respondents, may be more responsive to instrumental motives for crime than White respondents.

It has been hypothesized that males and females follow different pathways to adult crime. This line of reasoning has given rise to what is commonly referred to as the gendered pathways model or school of criminology (Chesney-Lind & Pasko, 2004). According to the gendered pathways to crime model, physical and sexual abuse, substance misuse, mental health difficulties, and relationship issues are more salient in the development of female offending than they are in the development of male offending, a supposition supported by a growing body of research and theoretical analysis (Jones, Brown, Wanamaker, & Greiner, 2014; Kruttschnitt, 2013; Salisbury & van Voorhis, 2009; Turanovic, Reisig, & Pratt, 2015). It is possible that many of these gender-responsive criminogenic influences could lead to feelings of low self-esteem, and while low self-esteem is not a criminogenic need in male offenders, there is speculation that it may be a criminogenic need in female offenders (Hubbard, 2006). This effect could be limited to White offenders, however, based on the results of one study which found that Black female delinquents registered significantly higher levels of self-esteem than White female delinquents (Holsinger & Holsinger, 2005). What this all means is that female offenders, at least White female offenders, may be more likely to adopt expressive motives for crime than male offenders.

From prior research (Walters et al., 2007), it is surmised that instrumental and expressive motives for crime may go hand-in-hand with proactive and reactive criminal thinking, respectively. Given the possibility that race may be moderated by gender or that gender may be moderated by race (Holsinger & Holsinger, 2005), both variables and their interaction need to be taken into account when conducting research on proactive and reactive criminal thinking. These fine-grained moderating effects were not part of the current hypotheses, however, because they were only suggested by the results of one study. Only the general countervailing relationships were predicted: that is, proactive criminal thinking should be higher in Black than White respondents and reactive criminal thinking should be higher in female than male respondents. Two hypotheses were accordingly tested in this study. The first hypothesis stated that Black male and female offenders on federal probation or supervised release would register significantly higher scores on the Proactive Criminal Thinking (PCT) scale of the PICTS compared with White male and female offenders. The second hypothesis held that Black and White female offenders on federal probation or supervised release would record significantly higher scores on the Reactive Criminal Thinking (RCT) scale of the PICTS than Black and White male offenders.

Method

Participants

In November 2009, the Federal Probation and Pretrial Services of the U.S. Courts began administering the PICTS routinely to all federal offenders under community supervision (i.e., probation and supervised release). The difference between federal probation and supervised release is that with probation, the judge sentences the individual directly to community supervision, whereas with supervised release, the individual is supervised in the community following release from prison. Approximately 80% of the individuals under postconviction supervision in the federal system are on supervised release. Hence, most of the participants in the current study had recently been released from federal prison.

Between 2009 and 2013, a total of 104,609 male and female Black and White (non-Hispanic) supervisees had completed the PICTS at least once. Only the first administration of the PICTS was considered for inclusion in this study. Two different samples of 2,000 supervisees each, an initial sample followed by a cross-validation sample, were selected for this study using disproportionate stratified random sampling (i.e., 500 White males, 500 Black males, 500 White females, and 500 Black females). One reason for creating smaller samples from the large original sample was that in a sample of 104,609 participants, even trivial effects are likely to be significant. A second reason for creating the smaller samples was to even out the number of males and females in the samples and allow for a fairer test of the second hypothesis. The use of the two samples permitted the surfacing of small but nontrivial effects, provided a better test of the second hypothesis because males and females were sampled in equal numbers, and allowed for cross-validation of initial results.

Measures

There were two independent variables included in the present study: race and gender. Race was coded as either 1 (White, non-Hispanic) or 2 (Black) and gender was coded as either 0 (male) or 1 (female). The two dependent variables were the PCT and RCT scales of the PICTS (Walters, 1995), a self-report measure designed to assess criminal thinking. Each item on the 80-item PICTS is rated on a 4-point scale (1 = disagree, 2 = uncertain, 3 = agree, 4 = strongly agree). The 32-item PCT and 24-item RCT were selected for use in the current investigation because proactive and reactive criminal thinking were the focus of the two hypotheses. The PICTS PCT and RCT scales demonstrate good internal consistency, test–retest reliability, and predictive validity (Walters, 2002, 2012). Two control variables were also included in the analyses: age at time of PICTS completion and supervision status, coded as probation (1) or supervised release (2). Given that supervised release participants had ordinarily been in prison prior to being placed on supervised release, the supervision status variable was considered a proxy for criminal history.

Research Design and Procedure

Two separate ANCOVAs were performed on each sample, one in which the raw PCT score served as the dependent variable and the other in which the raw RCT score served as the dependent variable. The independent variables in all four analyses were race and gender and the two control variables were age and supervision status. The Race × Gender interaction was also included in each analysis. In the event of a significant Race × Gender interaction, the simple main effects were examined. Ninety-six percent of the participants in the initial sample and 95.4% of the participants in the cross-validation sample had complete data on all 80 PICTS items. In situations where there were missing data on a scale, the raw score was pro-rated based on the scores from the non-missing items. The Administrative Office (AO) of the U.S. Courts released these data to the author for use in research and the Institutional Review Board at Kutztown University approved the use of these data for research purposes.

Results

Descriptive statistics and correlations for the six variables and two samples included in the current investigation are listed in Table 1. The zero-order correlations in both samples provided preliminary support for the two hypotheses tested in this study. First, race achieved a Bonferroni-corrected statistically significant correlation with the PCT (i.e., Blacks achieved significantly higher PCT scores than Whites) but gender did not. Second, gender achieved a Bonferroni-corrected statistically significant correlation with the RCT (i.e., females achieved significantly higher RCT scores than males), but the race-RCT correlation was nonsignificant.

Table 1.

Descriptive Statistics and Correlations for the Six Variables in Both Samples.

Variable	M	SD	Range	Status	Race	Gender	PCT	RCT
Initial sample (N = 2,000)
1. Age	39.64	11.28	18-87	.04	−.13*	−.12*	−.14*	−.11*
2. Status	1.79	0.41	1-2		.02	−.21*	.17*	.04
3. Race	1.50	0.50	1-2			.00	.11*	−.00
4. Gender	0.50	0.50	0-1				−.04	.10*
5. PCT	45.63	9.90	32-99					.66*
6. RCT	36.58	10.23	24-86
Cross-validation sample (N = 2,000)
1. Age	39.40	11.53	17-80	.05	−.20^†	−.08^†	−.16^†	−.12^†
2. Status	1.76	0.42	1-2		.02	−.23^†	.16^†	.07^†
3. Race	1.50	0.50	1-2			.00	.17^†	.04
4. Gender	0.50	0.50	0-1				.02	.10^†
5. PCT	45.62	10.02	32-112					.66^†
6. RCT	36.80	10.26	24-82

Note. Range = range of scores in current sample; status = supervision status (probation = 1, supervised release = 2); race = participant race (White = 1, Black = 2); gender = participant sex (male = 0, female = 1); PCT = PICTS Proactive Criminal Thinking scale; RCT = PICTS Reactive Criminal Thinking scale; age = age at time participant completed the PICTS; PICTS = Psychological Inventory of Criminal Thinking Styles.

*p < .0033 (Bonferroni-corrected alpha level = .05/15 correlations), initial sample.

†

p < .0033 (Bonferroni-corrected alpha level = .05/15 correlations), cross-validation sample.

Table 2 summarizes the results of the PCT analyses. In the initial sample, race but not gender was found to predict PCT scores and the Race × Gender interaction approached statistical significance (p = .06). A review of the simple main effects in the initial sample revealed that Black males achieved significantly higher mean PCT scores than White males, 47.12 versus 44.56, t(998) = 4.15, p < .001, but Black females did not achieve significantly higher PCT scores than White females, 45.86 versus 44.96, t(998) = 1.47, p = .14. In the cross-validation sample, both race and gender predicted the PCT and the Race × Gender interaction was statistically significant (see Figure 1). The simple main effects in the cross-validation sample indicated that Black males recorded significantly higher mean PCT scores than White males, 47.07 versus 43.10, t(998) = 6.38, p < .001, and Black females scored significantly higher on the PCT than White females, 46.94 versus 45.36, t(998) = 2.57, p < .05.

Table 2.

ANCOVA of Race and Gender as Predictors of Proactive Criminal Thinking in Both Samples.

Source	Initial sample					Cross-validation sample
Source	SS	df	MS	F	p	SS	df	MS	F	p
Intercept	113,749.01	1	113,749.90	1,229.97	<.001	121,888.79	1	121,888.79	1,316.67	<.001
Age	3,459.12	1	3,459.12	37.40	<.001	3,408.99	1	3,408.99	36.82	<.001
Status	4,987.72	1	4,987.72	53.93	<.001	5,745.33	1	5,745.33	62.06	<.001
Race	1,467.53	1	1,467.53	15.87	<.001	3,674.84	1	3,674.84	39.70	<.001
Gender	87.56	1	87.56	0.95	.331	533.02	1	533.02	5.76	.017
Race × Gender	341.77	1	341.77	3.70	.055	703.88	1	703.88	7.60	.006
Error	184,407.47	1994	92.48			184,591.56	1994	92.57

Note. N = 2,000 (initial sample) and 2,000 (cross-validation sample). Intercept = y intercept; age = age at time participant completed the PICTS; status = supervision status (probation = 1, supervised release = 2); race = participant race (White = 1, Black = 2); gender = participant sex (male = 0, female = 1); Race × Gender = interaction between race and gender; error = error term; dependent variable = PICTS Proactive Criminal Thinking scale; PICTS = Psychological Inventory of Criminal Thinking Styles; SS = sum of squares; df = degrees of freedom; MS = mean square; F = F-test results; p = significance leve of F-test.

Figure 1.

Effect of race and gender on the PICTS PCT score in the cross-validation sample (N = 2,000).

Table 3 summarizes the results of the RCT analyses. In the initial sample, gender but not race was found to predict RCT scores and the Race × Gender interaction approached statistical significance (p = .07). An analysis of the simple main effects in the initial sample showed that White females achieved significantly higher mean RCT scores than White males, 38.19 versus 35.34, t(998) = 4.39, p < .001, but Black females did not achieve significantly higher mean RCT scores than Black males, 36.99 versus 35.78, t(998) = 1.87, p = .06. Gender but not race also predicted RCT scores in the cross-validation sample, although this time, the Race × Gender interaction was significant (see Figure 2). Simple main effect analyses revealed that White females recorded significantly higher mean RCT scores than White males, 38.53 versus 34.82, t(998) = 4.98, p < .001, but Black females and males did not differ significantly on the RCT, 37.37 versus 36.47, t(998) = 1.38, p = .17.

Table 3.

ANCOVA of Race and Gender as Predictors of Reactive Criminal Thinking in Both Samples.

Source	Initial sample					Cross-validation sample
Source	SS	df	MS	F	p	SS	df	MS	F	p
Intercept	83,958.85	1	83,958.85	820.52	<.001	88,142.86	1	88,142.86	866.07	<.001
Age	1,916.39	1	1,916.39	18.73	<.001	2,210.64	1	2,210.64	21.72	<.001
Status	838.89	1	838.89	8.20	.004	1,721.98	1	1,721.98	16.92	<.001
Race	71.77	1	71.77	0.70	.402	28.60	1	28.60	0.28	.596
Gender	1,945.53	1	1,945.53	19.01	<.001	2,504.83	1	2,504.83	24.62	<.001
Race × Gender	330.97	1	330.97	3.24	.072	964.65	1	964.65	9.48	.002
Error	204,035.32	1,994	102.32			202,935.15	1,994	101.77

Note. N = 2,000 (initial sample) and 2,000 (cross-validation sample). Intercept = y intercept; age = age at time participant completed the PICTS; status = supervision status (probation = 1, supervised release = 2); Race = participant race (White = 1, Black = 2); gender = participant sex (male = 0, female = 1); Race × Gender = interaction between race and gender; error = error term; dependent variable = PICTS Reactive Criminal Thinking scale; PICTS = Psychological Inventory of Criminal Thinking Styles: SS = sum of squares; df = degrees of freedom; MS = mean square; F = F-test results; p = significance level of F-test.

Figure 2.

Effect of gender and race on the PICTS RCT score in the cross-validation sample (N = 2,000).

To evaluate the practical utility of the current results, a series of receiver operating characteristic (ROC) analyses were computed with race or sex as the state variable and the PCT or RCT as the test variable. ROC analysis entails plotting the true positive rate of the test variable as a function of its false negative rate to produce an area under the curve (AUC) accuracy estimate. The AUC indicates the probability that of two randomly selected cases, the one with the higher score will belong to the target group (Black in the PCT analysis and female in the RCT analysis). As indicated by the results summarized in Table 4, the total sample AUC results were statistically significant but of questionable utility for prediction (<.600). The refined analyses (restricting the PCT analysis to males and the RCT analysis to Whites), however, predicted outcome at a modest to moderate level.

Table 4.

Receiver Operating Characteristic (ROC) Analyses for the Full and Refined Complements of the Initial and Cross-Validation Samples.

	Initial sample				Cross-validation sample
	AUC	SE	95% CI	p	AUC	SE	95% CI	p
Full complement
PCT	.572	.013	[.547, .597]	<.001	.612	.012	[.587, .636]	<.001
RCT	.548	.013	[.523, .573]	<.001	.562	.013	[.537, .587]	<.001
Refined complement
PCT	.610	.018	[.576, .645]	<.001	.671	.017	[.638, .704]	<.001
RCT	.581	.018	[.546, .616]	<.001	.621	.018	[.586, .655]	<.001

Note. AUC = area under the curve; SE = standard error under the nonparametric assumption; 95% CI = asymptotic 95% confidence interval of the AUC; p = asymptotic significance level; full complement = analyses conducted on all 2,000 participants; PCT = race (Black) as the state variable and the Psychological Inventory of Criminal Thinking Styles (PICTS) Proactive Criminal Thinking scale as the test variable; RCT = gender (female) as the state variable and the PICTS Reactive Criminal Thinking scale as the test variable; refined complement = analyses conducted on a subset of 1,000 participants (PCT restricted to males and RCT restricted to Whites).

Discussion

The current results provide partial support for the two hypotheses tested in this study using two large samples of federal supervisees, the majority of whom had recently been released from federal prison. A more fine-tuned analysis of the results indicates that the proposed relationships between race and proactive criminal thinking and between gender and reactive criminal thinking are more circumscribed than originally thought, or at least they are substantially weaker at one level of the alternate independent variable than they are at the other level of the alternate independent variable. Hence, while consistent evidence of a race effect was obtained for proactive criminal thinking, with Blacks achieving significantly higher PCT scores than Whites, a significant effect across both samples was observed only in males. Likewise, a significant sex effect was identified for reactive criminal thinking, with females recording significantly higher RCT scores than males, but the effect was restricted to White participants. Bounded effects hinted at in previous research (Holsinger & Holsinger, 2005) were verified in the present study. Whereas all eight of the simple main effects were in the predicted direction, only five were statistically significant. Overall, then, the results offer circumscribed support for the proposed countervailing demographic–criminal thinking relationships proposed in the hypotheses, and using the current results to refine these relationships (i.e., restricting the Black–White difference in proactive criminal thinking to males and restricting the male–female difference in reactive criminal thinking to Whites) brings these proposed relationships into sharper focus.

The main point of this study was to advance theory—specifically, how we can make sense of the well-documented quantitative differences in offending between African Americans and Caucasian Americans (Gabbidon & Greene, 2013) and the equally well-documented qualitative differences in offending between males and females (Kruttschnitt, 2013). The significantly higher mean scores attained by African Americans on the PCT in comparison to Caucasian Americans and the modest-to-moderate accuracy of the PCT in classifying individual participants as Black or White in the male subsample provide support for Walters’s (2011) contention that African Americans, particularly African American men, are more motivated by the instrumental aspects of crime (i.e., crime as a means to an end) than Caucasian Americans. Structural and cultural factors likely contribute to Black–White differences in crime (Bellair & McNulty, 2005; Harer & Steffensmeier, 1992; Peterson & Krivo, 2005), but motivational factors, particularly the belief that crime can be viewed by those with limited resources as a way of getting ahead in America, cannot be ignored. Consequently, this supports Walters’s (2011) assertion that cultural, structural, and motivational factors combine to produce the quantitative differences between African Americans and Caucasian Americans that lead to the well-known Black–White differences in crime rates when official arrest and victimization data are evaluated (Steffensmeier, Feldmeyer, Harris, & Ulmer, 2011).

The qualitative differences observed between male and female offenders are the foundation of the gendered pathways theory of crime (Chesney-Lind & Pasko, 2004). Proponents of the gendered pathways model assume that males and females follow somewhat different pathways to crime and that female offending is qualitatively distinct from male offending. Whereas Black–White quantitative differences in crime are said to be motivated by the instrumental rewards of crime, male–female qualitative differences in crime are believed to be motivated by the expressive elements of crime. Hence, women are more likely to enter crime for emotional as opposed to material reasons and are more likely to recidivate because of family and relationship issues than men. Accordingly, reactive criminal thinking, which is heavily weighted with emotion, should play a larger role in female offending than it does in male offending. In the current study, reactive criminal thinking was significantly higher in White female offenders than in White male offenders, but there were no differences in reactive criminal thinking between Black male and female offenders. In a study comparing Black and White female delinquents, it was noted that White girls suffered more from early abuse, family conflict, mental health problems, and low self-esteem than Black girls, all features of the gendered pathways model (Holsinger & Holsinger, 2005). In the current research, expressive motives and reactive criminal thinking appeared to be restricted to White female offenders.

What needs to be kept in mind when interpreting the results of this study is that the mean differences in PICTS scores between groups were small. Even the largest mean group differences were less than half a standard deviation in size. This would seem to suggest that from a practical standpoint, the current results are limited, although the AUC estimates in the refined analyses (i.e., restricting the PCT analysis to males and the RCT analysis to females) indicated modest to moderate accuracy. One could use these results to argue that when working with Black male offenders, we should pay particular attention to proactive criminal thinking and instrumental motives for crime and target positive outcome expectancies for crime and moral decision-making, although these are issues that should be addressed with most offenders, not just Black male offenders. It could likewise be argued that when working with White female offenders, we should pay particular attention to reactive criminal thinking and expressive motives for crime and target self-control and problem-solving ability, but, again, these are issues that should be addressed with most offenders because of the extensive overlap that exists between proactive and reactive criminal thinking. Therefore, while there are practical implications to the current results, these implications are overshadowed by the theoretical implications discussed earlier.

Limitations

One of the principal limitations of this study is its uncertain generalizability. The current study was conducted on federal probationers and individuals on federal supervised release who ranged in age from 17 to 87. It is, therefore, unknown how well these results generalize to juvenile delinquents, state probationers and prisoners, and more serious and chronic federal offenders. A second limitation of this study is the relatively small size of the between-groups effects. Significant mean group differences were all under half a standard deviation. Despite the fact that these effects were small, they were far from trivial as evidenced by AUC classification accuracy results in the refined analyses. A third limitation of the present study is that it did not fully embrace Agnew’s (1995) recommendations for testing criminological theories. According to Agnew, a theory is most unique and therefore best tested by examining its motivational underpinnings rather than its independent variables. Of the four motivational processes he identified—moral evaluation of crime, rational evaluation of crime, negative affect, and freedom—two (moral evaluation of crime and freedom) relate closely to proactive and reactive criminal thinking, respectively. Each motivational process is considered an intervening variable in Agnew’s (1995) system, mediating the relationship between a theory’s independent and dependent (criminal outcome) variables. The current investigation, however, conceptualized the cognitive-motivational processes (proactive and reactive criminal thinking) as dependent variables rather than as mediator variables.

For proactive and reactive criminal thinking to be classified as intervening cognitive-motivational variables in a full analysis of a criminological theory, both would need to be treated as mediators and placed temporally between the demographic (independent) and criminal outcome (dependent) measures. The research design required to test a mediation model for conditional (moderator) effects is moderated mediation. In this particular design, the interactions between the independent variables (race and sex) and mediator variables (proactive and reactive criminal thinking) are computed and tested for significance. This represents Model 1 of the five models held by Preacher, Rucker, and Hayes (2007) to be capable of testing moderated mediation. Model 1 tests conditional indirect effects by making one or more independent variables (demographic measures) moderators of the mediator (PICTS scale). If one or more of these moderator effects is significant, then this would suggest that proactive and/or reactive criminal thinking are cognitive-motivational variables in a demographic-crime theory of offending consonant with the views of Agnew (1995). If none of these moderator effects is significant, then this would suggest that proactive and reactive criminal thinking do not operate as cognitive-motivational variables in a demographic-crime theory of offending. The current study was an initial effort to explain racial and gender differences in crime, but it needs to be followed up by research employing more sophisticated research designs, procedures, and analyses before a more definitive conclusion can be reached.

Footnotes

Acknowledgements

The author would like to thank the Administrative Office of the U.S. Courts for allowing access to these data.

Author’s Note

Glenn D. Walters is the author of the Psychological Inventory of Criminal Thinking Styles (PICTS) and has received remuneration from the sale of the manual in the past.

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) received no financial support for the research, authorship, and/or publication of this article.

Author Biography

Glenn D. Walters, PhD, is a full professor in the Department of Criminal Justice at Kutztown University. His current research interests include offender assessment, mediation analysis, and theoretical integration in criminology.

References

Agnew

(1995). Testing the leading crime theories: An alternative strategy focusing on motivational processes. Journal of Research in Crime & Delinquency, 32, 363-398.

Andrews

D. A.

Bonta

(2010). The psychology of criminal conduct (5th ed.). New Providence, NJ: Matthew Bender.

Bellair

P. E.

McNulty

T. L.

(2005). Beyond the bell curve: Community disadvantage and the explanation of black–white differences in adolescent violence. Criminology, 43, 1135-1168.

Bernard

T. J.

Snipes

J. B.

Gerould

A. L.

(2010). Vold’s theoretical criminology (6th ed.). New York, NY: Oxford University Press.

Chesney-Lind

Pasko

(2004). The female offender: Girls, women and crime (2nd ed.). Thousand Oaks, CA: Sage.

Cornish

D. B.

Clarke

R. V.

(Eds.). (1986). The reasoning criminal: Rational choice perspectives on offending. New York, NY: Springer-Verlag.

Gabbidon

S. L.

Greene

H. T.

(2013). Race and crime (3rd ed.). Thousand Oaks, CA: Sage.

Harer

M. D.

Steffensmeier

(1992). The differing effects of economic inequality on black and white rates of violence. Social Forces, 70, 1035-1054.

Harris

A. R.

Lewis

(1974, August). Race and criminal deviance: A study of youthful offenders. Paper Presented at the Annual Meeting of the American Sociological Association, San Francisco, CA.

10.

Holsinger

A. M.

(2005). Differential pathways to violence and self-injurious behavior: African-American and white girls in the juvenile justice system. Journal of Research on Crime & Delinquency, 42, 211-242.

11.

Hubbard

D. J.

(2006). Should we be targeting self-esteem in treatment for offenders: Do gender and race matter in whether self-esteem matters? Journal of Offender Rehabilitation, 44, 39-57.

12.

Jones

N. J.

Brown

S. L.

Wanamaker

K. A.

Greiner

L. E.

(2014). A quantitative exploration of gendered pathways to crime in a sample of male and female juvenile offenders. Feminist Criminology, 9, 113-136.

13.

Kohlberg

(1981). Essays on moral development, Vol. I: The philosophy of moral development. New York, NY: Harper & Row.

14.

Kruttschnitt

(2013). Gender and crime. Annual Review of Sociology, 39, 291-308.

15.

Owens-Sabir

M. C.

(2005). The continuum of control and self-esteem: A study of delinquency. Dissertation Abstracts International: Section A, 66(4), 1507.

16.

Peterson

R. D.

Krivo

L. J.

(2005). Macrostructural analyses of race, ethnicity, and violent crime: Recent lessons and new directions for research. Annual Review of Sociology, 31, 331-356.

17.

Preacher

K. J.

Rucker

D. D.

Hayes

A. F.

(2007). Addressing moderated mediation hypotheses: Theory, methods, and prescriptions. Multivariate Behavioral Research, 42, 185-227.

18.

Salisbury

E. J.

van Voorhis

(2009). Gendered pathways: A quantitative investigation of women probationers’ paths to incarceration. Criminal Justice and Behavior, 36, 541-566.

19.

Steffensmeier

Feldmeyer

Harris

C. T.

Ulmer

J. T.

(2011). Reassessing trends in black violent crime, 1980-2008: Sorting out the “Hispanic effect” in Uniform Crime Reports arrests, National Crime Victimization Survey offender estimates, and U.S. prisoner counts. Criminology, 49, 197-251.

20.

Sykes

G. M.

Matza

(1957). Techniques of neutralization: A theory of delinquency. American Sociological Review, 22, 664-670.

21.

Turanovic

Reisig

Pratt

(2015). Risky lifestyles, low self-control, and violent victimization across gendered pathways to crime. Journal of Quantitative Criminology, 31, 183-206.

22.

Walters

G. D.

(1990). The criminal lifestyle: Patterns of serious criminal conduct. Newbury Park, CA: Sage.

23.

Walters

G. D.

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

24.

Walters

G. D.

(2002). The Psychological Inventory of Criminal Thinking Styles (PICTS): A review and meta-analysis. Assessment, 9, 278-291.

25.

Walters

G. D.

(2007). Measuring proactive and reactive criminal thinking with the PICTS: Correlations with outcome expectancies and hostile attribution biases. Journal of Interpersonal Violence, 22, 371-385.

26.

Walters

G. D.

(2011). Black–white differences in positive outcome expectancies for crime: A study of male federal prison inmates. Journal of Criminal Justice, 39, 192-197.

27.

Walters

G. D.

(2012). Crime in a psychological context: From career criminals to criminal careers. Thousand Oaks, CA: Sage.

28.

Walters

G. D.

(2014). Relationships among race, education, criminal thinking, and recidivism: Moderator and mediator effects. Assessment, 21, 82-91.

29.

Walters

G. D.

(2015). A two-dimensional model of psychopathy and antisocial behavior: A multi-sample investigation using items from the Psychopathy Checklist-Revised. Personality and Individual Differences, 78, 88-93.

30.

Walters

G. D.

(2016). Friends, cognition, and delinquency: Proactive and reactive criminal thinking as mediators of the peer influence and peer selection effects among male delinquents. Justice Quarterly, 33, 1055-1079.

31.

Walters

G. D.

(in press). Proactive and reactive criminal thinking and self-reported offending: A cross-national survey of 7th thru 9th grade boys and girls. International Journal of Offender Therapy and Comparative Criminology.

32.

Walters

G. D.

Frederick

A. A.

Schlauch

(2007). Postdicting arrests for proactive and reactive aggression with the PICTS proactive and reactive composite scales. Journal of Interpersonal Violence, 22, 1415-1430.

33.

Yochelson

Samenow

S. E.

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