2

Exposure to Criminal Environment and Criminal Social Identity in a Sample of Adult Prisoners: The Moderating Role of Psychopathic Traits

Nicole Sherretts University of Huddersfield

Daniel Boduszek University of Huddersfield and SWPS University of Social

Sciences and Humanities

Agata Debowska University of Chester

The purpose of this study was to investigate the role of period of incarceration, criminal friend index (a retrospective measure intended to quantify criminal associations before 1st incarceration), and 4 psy- chopathy factors (interpersonal manipulation, callous affect, erratic lifestyle, and antisocial behavior) in criminal social identity (CSI) while controlling for age and gender. Participants were a sample of 501 incarcerated offenders (male n � 293; female n � 208) from 3 prisons located in Pennsylvania State. Moderated regression analyses indicated no significant direct association between period of incarceration and CSI or between criminal friend index and Measure of Criminal Social Identity (MCSI). However, a significant moderating effect of interpersonal manipulation on the relationship between period of incarceration and MCSI was observed. Period of incarceration was significantly positively correlated with MCSI (particularly with the in-group ties subscale) for only those offenders who scored high (1 SD above the mean) on interpersonal manipulation and significantly negatively correlated for those who scored low (1 SD below the mean) on interpersonal manipulation. Also, criminal friend index was positively significantly associated with in-group ties for high levels (1 SD above the mean) of callous affect. The main findings provide evidence for the claim that prisoners are likely to simulate changes in identity through the formation of bonds with other offenders and that this can be achieved using interpersonal manipulation skills.

Keywords: criminal social identity, psychopathic traits, incarcerated offenders, moderated regression analysis

Identity is composed of meanings that individuals assign to the roles they play in different social contexts and can be renegotiated in the process of interaction with others as and when needed (Stryker & Burke, 2000). Turner (1982) distinguished between two types of identities: personal and social. Personal identity accentu- ates individuals’ uniqueness and is most resistant to change. Be- havior that arises as a function of personal identity is largely guided by psychological variables. Social identity, on the other hand, is determined by category-based processes, stresses individ- uals’ similarities with the reference group, and positions them “within structured social arrangements” (Vryan, Adler, & Adler, 2003, p. 371).

Social identity theory (SIT; Tajfel & Turner, 1979) is concerned with the latter and was developed to explain prejudice, discrimi-

nation, and behavior in intergroup relations (Hogg & Reid, 2006). SIT posits that people strive to attain a positive social identity in order to protect their self-esteem and that this can be achieved through favorable comparisons between in-group and out-group members, a process referred to as in-group favoritism (Tajfel & Turner, 1979). Because positive social identity is the source of self-enhancement, in the event of an unsatisfactory outcome of such an evaluation, individuals may choose to leave their reference group or distance themselves from its members (Hogg & Reid, 2006). Another conceptual framework, self-categorization theory (SCT; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987), derived from SIT, focuses on social– cognitive processes that influence people’s identification with certain groups, the propensity toward construing themselves in group terms, and the readiness to mani- fest group behaviors. SCT considers self-categorizations to be central in the formation of both personal and social identities. To elaborate, people tend to classify themselves and others in certain social categories, such as gender, nationality, or religious affilia- tion, which provides an important cognitive structure helpful in ordering the social environment. The process of categorization affects their self-concept, feelings, and behaviors (Hogg & Reid, 2006).

The concepts of SIT and SCT have formed the basis for the development of criminal social identity (CSI) theory (Boduszek & Hyland, 2011) and its updated version, the integrated psychosocial

This article was published Online First April 7, 2016. Nicole Sherretts, Department of Behavioural and Social Sciences, Uni-

versity of Huddersfield; Daniel Boduszek, Department of Behavioural and Social Sciences, University of Huddersfield, and Psychology Department, SWPS University of Social Sciences and Humanities; Agata Debowska, Department of Psychology, University of Chester.

Correspondence concerning this article should be addressed to Daniel Boduszek, Department of Behavioural and Social Sciences, University of Huddersfield, Queensgate, Huddersfield HD1 3DH, United King- dom. E-mail: [email protected]

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

Law and Human Behavior © 2016 American Psychological Association 2016, Vol. 40, No. 4, 430 – 439 0147-7307/16/$12.00 http://dx.doi.org/10.1037/lhb0000188

430

model of criminal social identity (IPM-CSI; Boduszek, Dhingra, & Debowska, in press), which explain the etiology and consequences of identity within a specific social group. On the basis of Camer- on’s (2004) earlier research into the factor structure of the measure of social identity, the model of CSI was proposed to be composed of three dimensions, namely cognitive centrality, in-group affect, and in-group ties (Boduszek, Adamson, Shevlin, & Hyland, 2012a). Cognitive centrality stresses the cognitive importance of belonging to a particular group. Criminal identity for individuals scoring high on this aspect of CSI is seen as central to their self-concept, which renders them more likely to endorse the group norms and act accordingly, even in the absence of other group members. In-group affect refers to the positive emotional valence of belonging to a criminal group and is argued to be developed to reduce anxiety associated with the discrepancy between ideal self and actual self. The final factor, in-group ties, pertains to the psychological perception of resemblance and emotional connec- tion with other members of a particular group. Individuals with strong in-group ties are persistently readier to display behaviors condoned by the group in order to demonstrate their conformity (Boduszek, Adamson, Shevlin, & Hyland, 2012a; Boduszek, Ad- amson, Shevlin, Mallett, & Hyland, 2013; Boduszek, O’Shea, Dhingra, & Hyland, 2014).

Sociopsychological processes that influence the emergence of criminal identity include associations with criminal friends, per- ceptions of self-esteem, and childhood experiences (e.g., peer rejection and dysfunctional family environment; Boduszek, Ad- amson, Shevlin, Mallett, & Hyland, 2013; Boduszek & Hyland, 2011). Moreover, the higher the degree of identification with antisocial friends, the greater the likelihood of developing criminal cognitive structures and engaging in criminal behavior (Boduszek & Hyland, 2012; Boduszek, Hyland, Pedziszczak, & Kielkiewicz, 2012; Holsinger, 1999; Simourd, 1997; Simourd & Van De Ven, 1999). To date, research in the area of CSI has been conducted with male samples only; however, some studies have suggested that women are more likely to form stronger bonds and identification with in-group members because of the greater need to belong (e.g., B. B. Brown & Lohr, 1987; R. Brown, Condor, Mathews, Wade, & Williams, 1986; Kiesner, Cadinu, Poulin, & Bucci, 2002; Newman, Lohman, & Newman, 2007). Women were also reported to have wider social networks from which to garner help and encouragement (Friborg, Barlaug, Martinussen, Rosenvinge, & Hjemdal, 2005; Fri- borg, Hjemdal, Rosenvinge, & Martinussen, 2003). As such, it ap- pears that female prisoners may be more susceptible to group social- ization processes.

CSI was also found to have important psycho-social and mental health implications. Specifically, criminal identity was reported to predict criminal thinking style (Boduszek, Adamson, Shevlin, Hy- land, & Bourke, 2013; Boduszek, Shevlin, Adamson, & Hyland, 2013). Cognitive centrality correlated significantly with violent criminal behavior, whereas increased in-group affect was associ- ated with nonviolent criminal behavior (Boduszek, Hyland, et al., 2013). Using a sample of 415 incarcerated juvenile offenders, Shagufta, Boduszek, Dhingra, and Kola-Palmer (2015) found high in-group affect to serve as a protective factor against suicide ideation.

It has been indicated that environmental factors, such as being subject to incarceration, can affect individuals’ cognitive processes (Clemmer, 1940). For example, Rhodes (1979) found that incar-

cerated offenders, due to the constant exposure to other prisoners, tend to develop deviant attitudes while serving their sentence. In another study with Polish male prisoners and a sample of men drawn from the general population, a positive significant effect of imprisonment on cognitive distortions pertaining to rape and rape victims was found (Debowska, Boduszek, Dhingra, & DeLisi, in press). Consequently, prisoners seem to be subject to group so- cialization processes akin to those found among community-based groups and may develop an identification with the group’s values. It is important to note that because SIT predicts that, prior to the acquisition of group attitudes, social identity must be formed (Boduszek, Adamson, Shevlin, Hyland, & Bourke, 2013; Tajfel & Turner, 1979), a similar effect of incarceration on CSI can be expected. Indeed, Boduszek et al. (2014) found a significant effect of the number of arrests on CSI within a sample of male recidi- vistic offenders. A more-recent empirical investigation among 126 male juvenile offenders incarcerated in prisons in Khyber Pakh- tunkhwa, Pakistan, aimed to examine how primary psychopathy (incorporating affective and interpersonal features, as indexed by the Levenson Self-Report Psychopathy Scale (Levenson, Kiehl, & Fitzpatrick, 1995) may interact with period of confinement to predict criminal social identity (CSI) scores, while controlling for covariates (Boduszek, Dhingra, & Debowska, 2016). The results demonstrated that period of confinement had a significant positive effect on the formation of criminal identity but only for those participants who scored higher (1 SD above the mean) on primary psychopathy.

Indeed, it appears that personality traits should be examined in relation to the formation of criminal identity. Using a sample of Irish ex-prisoners, Boduszek, McLaughlin, and Hyland (2011) found psychoticism to be a strong predictor of criminal cognitions. Two personality traits (Extraversion and Psychoticism) were also reported to form significant associations with criminal thinking style, a construct strongly related with criminal social identity (Boduszek, Adamson, Shevlin, & Hyland, 2012b). Of importance, the IPM-CSI (Boduszek et al. in press) posits that exposure to criminal environment before first incarceration (such as having criminal friends) may affect the development of CSI, especially in the presence of certain personality moderators (e.g., psychopathic traits). Therefore, given the theoretical framework and previous empirical findings, the paucity of research into the role of person- ality traits in CSI appears to be an important omission that ought to be addressed.

Furthermore, although identities can be reconstructed, people are motivated to keep their self-conceptions stable in order to maintain harmony (Weigert & Gecas, 2003). For that purpose, they are likely to employ the strategy of selective affiliation, that is, interacting with similar others (Swann, 1987). However, selective affiliation is not available in all social contexts, for example in prison settings, where membership is not voluntary. Consequently, should prolonged discrepancies between self-concept and environ- ment appear, new self-relevant meanings can be created, which leads to identity change (Burke, 2006). Nevertheless, it is impor- tant to note here that identity change due to social adaptation is not simply a passive response to environmental stimuli (Barker, 2005). Instead, people are likely to manipulate this change by recognizing what they want, establishing a goal, and deciding on an appropriate course of action that would bring them closer to the desired object (Blumer, 1966). In line with Tajfel and Turner’s (1979) assump-

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

431PSYCHOPATHY, PRISON, AND CRIMINAL SOCIAL IDENTITY

tion that positive social identity is necessary for self-enhancement, Goffman (1963, 1990) argued that the self manages impressions of others and attempts at self-presentation that elicits positive evalu- ations. In light of this argument, it appears that criminal identity can be developed if categorizing self as a part of criminal group is seen as advantageous. On the basis of research findings by Bo- duszek et al. (2016), it may also be suggested that those more skilled at interpersonal manipulation will be more likely to adapt to social norms provided by prison settings and develop a criminal identity. Even though the researchers considered interpersonal (e.g., grandiosity, deceitfulness, and superficial charm) and affec- tive (e.g., low empathy, lack of remorse, emotional shallowness, and a failure to accept personal responsibility) features of psy- chopathy as a single dimension, other empirical studies have demonstrated that those traits correlate differently with external variables, such as reactive aggression and rape myth acceptance, as well as self-injurious thoughts and behavior (e.g., Debowska, Boduszek, Kola, & Hyland, 2014; Debowska, Mattison, & Bo- duszek, 2015; Debowska & Zeyrek Rios, 2015; Dhingra, Bo- duszek, Palmer, & Shevlin, 2015). Accordingly, in order to test the hypothesis that interpersonal manipulation and callous affect may be significantly related to the formation of criminal identity, future research in the area of CSI should include interpersonal and affective dimensions of psychopathy as separate components.

Current Study

Given the evidence that interaction between incarceration and psychopathic traits may lead to the intensification of CSI within a youth offending sample from Pakistan (Boduszek et al., 2016), it appears crucial that the concept be further examined within the adult Western prison population. Although research in the area is growing, still little is understood about the psycho-social factors having an impact on CSI in prison populations. Further, there is a lack of research looking at personality traits in relation to social identity in prison contexts. Thus, in the current study, we examine the effect of period of incarceration, criminal friend index (a retrospective measure intended to quantify criminal associations before first incarceration), and four psychopathy facets (interper- sonal manipulation, callous affect, erratic lifestyle, and antisocial behavior) on CSI while controlling for gender and age. On the basis of theory and previous studies, we expected to find a signif- icant positive effect of criminal friends and period of incarceration on CSI; however, we also predicted that the associations would be moderated by psychopathic traits.

Method

Participants

The opportunistic sample included 501 offenders incarcerated in three prisons in Pennsylvania State (men in maximum-security prison n � 131, men in medium-security prison n � 161, and women in maximum-security prison n � 208). The sample con- sisted of 141 first-time offenders, 266 repeated offenders, 80 with a life sentence, and 14 on death row. In the final analysis, the sample size was further reduced to 458 after pairwise deletion of missing data based on Little’s missing completely at random test, �2(234) � 254.40, p � .17. The respondents ranged in age from 19

to 76 (M � 39.53, SD � 11.79). Most offenders (56.8%) came from urban areas. Participants were Caucasian (55.7%), African American (26.1%), Hispanic (5.5%), or Other (12.7%). The fre- quency of imprisonment reported by offenders ranged from one to 22 times (M � 3.76, SD � 4.97).

Procedure

Ethical approval was granted by the Pennsylvania Department of Corrections review board. Authors in collaboration with the Prison Service selected three prisons for participation in the study (male maximum-security prison, male medium-security prison, and female maximum-security prison). Surveys were posted to all blocks within selected prisons and distributed to inmates using the opportunistic method. Given inmates’ standing as a vulnerable population and the potential that they may feel compelled to participate, it was made clear both in the consent form and verbally (by the prison personnel) that participation was voluntary without any form of reward. In addition, inmates were informed that they should not participate in the study if they could not read; however, they did not have to inform data collectors of the specific reason for not participating in the study. Inmates consenting to participate were told that all information they provided in this study was anonymous and that they could withdraw from the study at any time until the survey submission.

Materials

Measure of Criminal Social Identity (MCSI; Boduszek et al. 2012a). The MCSI consists of eight items and is based on Cameron’s (2004) Three-Dimensional Strength of Group Identifi- cation Scale. Each item is scored on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Scores range from 8 to 40, with higher scores indicating higher levels of criminal social identity. The scale is composed of three subscales: The cognitive centrality (three items) subscale measures the psychological salience of a criminal’s group identity; the in-group affect (two items) subscale measures a criminal’s felt attitude toward other in-group criminals; and the in-group ties (three items) subscale measures the level of personal bonding with other criminals. In the present sample, Cron- bach’s alphas were all acceptable (cognitive centrality � .69, in-group affect � .71, and in-group ties � .72).

Self-Report Psychopathy Scale—Short Form (SRP–SF; Paulhus, Neumann, & Hare, in press). The SRP–SF was used to assess self-reported psychopathic traits. The original SRP–III, generated on the basis of the Psychopathy Checklist—Revised (PCL–R; Hare, 1991), is a 64-item measure that yields a total score as well as four subscale scores. Paulhus et al. (in press) developed the shortened, 29-item form of the scale (SRP–SF) in order to reduce the administration time. For the purpose of the current study, the abbreviated version of the measure was used. The scale consists of four subscales: interpersonal manipulation (IPM; � � .78), callous affect (CA; � � .70), erratic lifestyle (ELS; � � .72), and antisocial behavior (ASB; � � .68). Items are scored on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Scores for the IPM, CA, and ELS subscales range from 7 to 35 and the ASB subscale from 8 to 40, with higher scores reflecting higher levels of psychopathic traits. Previous factor analytic work revealed a good fit for the four-factor model of the

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

432 SHERRETTS, BODUSZEK, AND DEBOWSKA

SRP–SF (e.g., Declercq, Carter, & Neumann, 2015; León-Mayer, Folino, Neumann, & Hare, 2015).

Criminal friend index (CFI; Mills & Kroner, 1999). The CFI is a measure intended to quantify criminal associations. Re- spondents are asked to recall three friends they spent most of their free time with before first incarceration and rate the amount of time spent with each one on the following scale: 1 � 0%–25%, 2 � 25%–50%, 3 � 50%–75%, and 4 � 75%–100%. Addition- ally, participants are asked to answer four questions in relation to the degree of criminal involvement of their friends: (1) Has this person ever committed a crime? (2) Does this person have a criminal record? (3) Has this person ever been to prison? and (4) Has this person tried to involve you in a crime? This measure is calculated by first taking the number associated with the percent- age of time spent with a particular friend and multiplying it by the number of yes responses to the four questions concerning that person’s criminal involvement. This is done for each of the three friends, and the resulting scores are added together to produce the criminal friend index (possible range of scores is from 0 to 48).

Analysis

Descriptive statistics, Pearson product–moment correlation co- efficients, and regression analysis were calculated using SPSS 22. Hierarchical moderated multiple regression analyses, as the rec- ommended method for testing interaction effects (Cohen & Cohen, 1983), were applied in order to investigate the moderating role of four psychopathy factors in the relationship between time in prison and CSI total score and criminal friend index and MCSI total score (as well as three dimensions of MCSI: cognitive centrality, in- group affect, and in-group ties), while controlling for gender of offenders and age. Simple slopes for the relationship between time in prison and CSI, as well as criminal friend index and MCSI, were investigated for low (1 SD below the mean), medium (mean), and high (1 SD above the mean) levels of psychopathic traits (inter- personal manipulation) using ModGraph 3.0 (Jose, 2013). Only standardized solution was reported.

Results

Descriptive Statistics and Correlations

Bivariate correlations and descriptive statistics including means and standard deviations for interpersonal manipulation, callous affect, erratic lifestyle, antisocial behavior, time in prison, criminal friends, cognitive centrality, in-group affect, in-group ties, and age are pre- sented in Table 1. Prisoners in the current sample revealed moderate levels of criminal social identity and psychopathy, with the highest scores on the antisocial behavior and erratic lifestyle dimensions of psychopathy.

Moderated Regression Analysis With MCSI as Outcome Variable

Hierarchical moderated regression analysis was performed to investigate the moderating effect of four psychopathy dimensions scores on the relationship between time spent in prison and MCSI total score, as well as criminal friend index and MCSI total score, while controlling for gender of offenders and age. Preliminary T

ab le

1 D

es cr

ip ti

ve S ta

ti st

ic s

(M a n d

S D

) a n d

C o rr

el a ti

o n s

(r W

it h

A ss

o ci

a te

d 9 5 %

C o n fi

d en

ce In

te rv

a ls

) fo

r A

ll C

o n ti

n u o u s

V a ri

a b le

s

V ar

ia bl

e 1

2 3

4 5

6 7

8 9

10

1. E

rr at

ic li

fe st

yl e

— 2.

A nt

is oc

ia l

be ha

vi or

.5 9�

� �

[. 53

,. 65

] —

3. C

al lo

us af

fe ct

.5 8�

� �

[. 52

,. 64

] .5

1� �

� [.

44 ,.

58 ]

— 4.

In te

rp er

so na

l m

an ip

ul at

io n

.5 8�

� �

[. 52

,. 64

] .5

7� �

� [.

51 ,.

63 ]

.6 2�

� �

[. 56

,. 67

] —

5. T

im e

in pr

is on

[i n

m on

th s]

.0 9

[� .0

1, .1

8] .1

4� �

[. 05

,. 23

] �

.0 2

[� .1

1, .0

7] �

.0 2

[� .1

1, .0

7] —

6. C

ri m

in al

fr ie

nd in

de x

.3 1�

� �

[. 23

,. 39

] .3

7� �

� [.

29 ,.

45 ]

.2 2�

� �

[. 13

,. 31

] .2

7� �

� [.

18 ,.

35 ]

� .0

3 [�

.1 2,

.0 6]

— 7.

C og

ni ti

ve ce

nt ra

li ty

.0 6

[� 03

,. 15

] .1

4� �

[. 05

,. 23

] .0

1 [�

.0 8,

.1 0]

.1 0�

[. 01

,. 19

] �

.0 1

[� .1

0, .0

8] .0

6 [�

.0 3,

.1 5]

— 8.

In -g

ro up

af fe

ct .1

1� [.

02 ,.

20 ]

.2 2�

� �

[. 13

,. 31

] .1

6� �

� [.

07 ,.

25 ]

.1 9�

� �

[. 10

,. 28

] .1

1� [.

02 ,.

20 ]

.0 3

[� .0

6, .1

2] .0

4 [�

.0 5,

.1 3]

— 9.

In -g

ro up

ti es

.4 0�

� �

[. 32

,. 47

] .3

6� �

� [.

28 ,.

44 ]

.2 6�

� �

[. 17

,. 34

] .3

7� �

� [.

29 ,.

45 ]

� .1

0� [�

.1 9,

� .0

1] .2

8� �

� [.

19 ,.

36 ]

.1 2�

[. 03

,. 21

] .2

9� �

� [.

20 ,.

37 ]

— 10

. A

ge �

.3 1�

� �

[� .3

9, �

.2 3]

� .1

5� �

� [�

.2 4,

� .0

6] �

.1 8�

� �

[� .2

7, �

.0 9]

� .1

0� [�

.1 9,

� .0

1] .5

0� �

� [.

43 ,.

57 ]

� .2

9� �

� [�

.3 7,

� .2

0] .0

3 [�

.0 6,

.1 2]

.0 8

[� .0

1, .1

7] �

.2 4�

� �

[� .3

2, �

.1 5]

— M

18 .0

9 20

.9 1

14 .9

5 14

.1 8

12 2.

41 11

.4 5

7. 91

2. 59

7. 40

39 .5

3 S D

5. 42

5. 82

4. 58

4. 90

10 9.

09 11

.2 4

2. 55

1. 25

2. 63

11 .7

9

� p

� .0

5. � �

p �

.0 1.

� � �

p �

.0 01

.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

433PSYCHOPATHY, PRISON, AND CRIMINAL SOCIAL IDENTITY

analyses revealed no violation of the assumptions of normality, linearity, multicollinearity, and homoscedasticity.

In the first step of the analysis, six predictors were entered: time spent in prison, criminal friend index, erratic lifestyle, antisocial behavior, callous affect, and interpersonal manipulation. This model (Model 1) was statistically significant, F(6, 421) � 15.39, p � .001, and explained 18% (R2 � .18) of variance in MCSI. Two psychop- athy facets—antisocial behavior and interpersonal manipulation— made a significant unique contribution to the model (see Table 2). However, no significant relationship between either total time spent in prison and MCSI or criminal friend index and MCSI was observed.

The second step consisted of entering interaction terms, coding the interaction between time spent in prison and four psychopathy factors, and criminal friend index and four psychopathy factors, while con- trolling for gender of offenders and age. After the interaction terms and covariates were entered, an additional 4% of variance in MCSI was explained (�R2 � .04, p � .038), and the final model (Model 2) as a whole explained 22% of variance in MCSI (R2 � .22), F(16, 411) � 7.12, p � .001. Just as in Model 1, antisocial behavior and interpersonal manipulation formed statistically significant direct rela- tionships with MCSI. Gender of offenders was negatively directly correlated with MCSI scores, suggesting that men were significantly less likely to report MCSI. It is important to note that there was no significant direct relationship between time in prison and MCSI scores. However, the relationship between interaction term (Time in Prison � Interpersonal Manipulation) and MCSI was statistically

significant, indicating that the effect of time spent in prison on MCSI depends on the level of interpersonal manipulation psychopathy factor.

In order to investigate this moderating effect further, we investi- gated simple slopes for the relationship between time in prison and MCSI for low (1 SD below the mean), medium (mean), and high (1 SD above the mean) levels of interpersonal manipulation (see Bate, Boduszek, Dhingra, & Bale, 2014; Boduszek et al. 2012b; Cohen & Cohen, 1983). For a graphical representation of moderating effects see Figure 1. Time in prison was positively significantly associated with MCSI for high levels ( 1 SD) of interpersonal manipulation ( � .18, SE � .09, 95% confidence interval [CI � 0.09, 0.27], p � .05). A negative significant association between time in prison and MCSI was found for low levels (�1 SD) of interpersonal manipula- tion ( � �.20, SE � .09, 95% CI [�0.21, �0.11], p � .05). The association between time in prison and MCSI for medium (mean) levels of interpersonal manipulation was negative yet statistically nonsignificant ( � �.01, SE � .06, 95% CI [�0.11, 0.09], p � .05). Therefore, the results suggest that the relationship between time spent in prison and MCSI depends on levels of interpersonal manipulation while controlling for other covariates in the model.

Moderated Regression Analyses With Three Subscales of MCSI as Outcome Variables

Given previous research indicating that three MCSI facets can form differential associations with external factors (Boduszek et al.

Table 2 Moderated Regression Analyses ( With 95% Confidence Intervals) With MCSI Total Score and Three Subscales (Cognitive Centrality, In-Group Affect, and In-Group Ties) as Outcome Variables

Model and variable Criminal social

identity Cognitive centrality In-group affect In-group ties

Model 1 Erratic lifestyle (ELS) .09 [�.03, .21] �.03 [�.16, .11] �.07 [�.20, .06] .20��� [.08, .32] Antisocial behavior (ASB) .22��� [.09, .34] .16� [.03, .29] .17�� [.05, .30] .14� [.03, .26] Callous affect (CA) �.11 [�.22, .01] �.12 [�.24, .01] .05 [�.07, .18] �.08 [�.19, .04] Interpersonal manipulation

(IPM) .22��� [.10, .34] .10 [�.03, .23] .12 [�.01, .25] .19�� [.07, .31] Time in prison (TP) �.06 [�.15, .03] �.04 [�.13, .06] .08 [�.02, .17] �.09 [�.18, .01] Criminal friend index

(CFI) .07 [�.02, .17] .01 [�.10, .11] �.05 [�.15, .05] .11 [�.01, .23]

Model 2 ELS .08 [�.05, .20] �.01 [�.15, .12] �.07 [�.20, .07] .17�� [.05, .29] ASB .21��� [.08, .33] .14� [.01, .28] .18�� [.05, .32] .14� [.02, .26] CA �.10 [�.22, .03] �.13 [�.26, .01] .05 [�.07, .18] �.06 [�.17, .06] IPM .23��� [.10, .36] .11 [�.04, .25] .14� [.01, .27] .19�� [.07, .32] TP �.01 [�.13, .11] �.02 [�.15, .11] .01 [�.11, .13] �.01 [�.11, .11] CFI .05 [�.05, .15] .01 [�.10, .11] �.03 [�.14, .07] .08 [�.01, .18] TP � ELS �.01 [�.14, .12] .09 [�.04, .23] �.13 [�.27, .01] �.08 [�.20, .04] TP � ASB .07 [�.06, .19] .03 [�.10, .16] .08 [�.05, .20] .04 [�.08, .15] TP � CA �.12 [�.26, .02] �.11 [�.25, .04] �.05 [�.20, .09] �.06 [�.19, .07] TP � IPM .19�� [.05, .33] .09 [�.06, .24] .14 [�.01, .29] .16� [.03, .29] CFI � ELS .08 [�.04, .21] .05 [�.08, .18] �.07 [�.20, .06] .12 [�.01, .23] CFI � ASB .03 [�.10, .15] .09 [�.04, .22] �.03 [�.16, .10] �.04 [�.16, .07] CFI � CA �.05 [�.17, .07] .04 [�.09, .17] .07 [�.05, .20] �.14� [�.25, �.03] CFI � IPM .04 [�.08, .15] �.06 [�.19, .06] .01 [�.11, .13] .11 [�.01, .22] Age .02 [�.09, .13] .06 [�.06, .18] .08 [�.03, .20] �.09 [�.20, .01] Gender (male � 1) �.24� [�.43, �.05] �.10 [�.31, .11] .02 [�.19, .22] �.21� [�.39, �.03]

� p � .05. �� p � .01. ��� p � .001.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

434 SHERRETTS, BODUSZEK, AND DEBOWSKA

2012b), additional moderated regression analyses using the same predictor variables, interaction terms, and covariates were per- formed for three MCSI subscales separately (see Table 2, last three columns). In the first step of hierarchical multiple regression, the main effect of four psychopathy dimensions, time in prison and criminal friend index on cognitive centrality, in-group affect, and in-group ties were examined. All these analyses (see Models 1 and 2 in Table 2) were statistically significant: in-group affect (R2 � .07), F(6, 433) � 5.26, p � .001; in-group ties (R2 � .23), F(6, 433) � 21.03, p � .001; and cognitive centrality (R2 � .03), F(6, 428) � 2.16, p � .04. Antisocial behavior was found to be a significant predictor of all three outcome variables. Additionally, erratic life- style and interpersonal manipulation were positively correlated with in-group ties.

In the second step of the analyses, interaction terms between time in prison and four psychopathy facets, and criminal friend index and four psychopathy facets, were entered while control- ling for covariates. This model was statistically nonsignificant for cognitive centrality, F(16, 418) � 1.71, p � .07, and explained 6% of variance (R2 � .06, �R2 � .03, p � .15). Antisocial behavior remained the only significant predictor of cognitive centrality. Model 2 for in-group affect was signifi- cant, F(16, 423) � 2.65, p � .001, and explained 9% of variance (R2 � .09); however, adding interaction terms and controlling for covariates did not contribute significantly to the model (�R2 � .02, p � .38). Two psychopathy dimensions, antisocial behavior and interpersonal manipulation, were sig- nificantly directly associated with in-group affect. Model 2 was also significant for in-group ties, F(16, 423) � 9.75, p � .001, and explained 27% of variance (R2 � .27, �R2 � .04, p � .01). Erratic lifestyle, antisocial behavior, and interpersonal manip- ulation were positively directly correlated with in-group ties. There was a significant negative association between gender of offenders and in-group ties. No significant direct relationship between either time in prison and in-group ties and criminal friend index and MCSI was established; however, the relation-

ship between interaction term (Time in Prison � Interpersonal Manipulation) and in-group ties was statistically significant. This suggested that the relationship between time spent in prison and in-group ties depends on the level of interpersonal manipulation psychopathy factor. Simple slopes for this rela- tionship were investigated for low (1 SD below the mean), medium (mean), and high (1 SD above the mean) levels of interpersonal manipulation (see Figure 2). Time in prison was positively significantly associated with in-group ties for high levels ( 1 SD) of interpersonal manipulation ( � .15, SE � .09, 95% CI [0.06, 0.24], p � .05). The simple slope for low levels (�1 SD) of interpersonal manipulation indicated a neg- ative significant association between time in prison and in- group ties ( � �.17, SE � .09, 95% CI [�0.26, �0.08], p � .05). There was a nonsignificant association between time in prison and in-group ties for medium (mean) levels of interper- sonal manipulation ( � �.01, SE � .05, 95% CI [�0.1, 0.09], p � .05).

Moreover, the relationship between interaction term (Crimi- nal Friend Index � Callous Affect) and in-group ties was statistically significant. This suggested that the relationship between criminal friend index and in-group ties depends on the level of callous affect traits. Simple slopes for this relationship were investigated for low (1 SD below the mean), medium (mean), and high (1 SD above the mean) levels of callous affect (see Figure 3). Criminal friend index was positively signifi- cantly associated with in-group ties for high levels ( 1 SD) of callous affect ( � .24, SE � .07, 95% CI [0.15, 0.33], p � .001). The simple slope for low levels (�1 SD) of callous affect indicated a negative nonsignificant association between crimi- nal friend index and in-group ties ( � �.08, SE � .07, 95% CI [�0.17, 0.02], p � .05). There was a nonsignificant association between criminal friend index and in-group ties for medium (mean) levels of callous affect ( � .08, SE � .04, 95% CI [�0.02, 0.17], p � .05).

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

low med high

M C

S I

Total time spent in prison

Figure 1. The moderating role of interpersonal manipulation in the relationship between time spent in prison and MCSI total score. The solid line with square markers indicates high ( 1 SD) interpersonal manipulation, the dotted line with triangle markers indicates medium (mean) interpersonal manipulation, and the dashed line with cross markers indicates low (�1 SD) interpersonal manipulation. med � medium.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

435PSYCHOPATHY, PRISON, AND CRIMINAL SOCIAL IDENTITY

Discussion

Few studies with sound methodological designs have examined the influence of the period of incarceration, criminal friend index, and psychopathic traits on the amplification of criminal identity. Additionally, to date no known study has explored the influence of different psycho-social processes on the three MCSI components (cognitive centrality, in-group affect, and in-group ties). Further, most previous empirical investigations have utilized male samples only. The main purpose of the present study, therefore, was to address those limitations by examining the role of the period of incarceration and psychopathic traits in MCSI total score and three MCSI dimensions, while controlling for gender and age.

Boduszek et al. (2014) found the number of arrests to be significantly positively associated with MCSI. Additionally, pre- vious research demonstrated a significant effect of imprisonment on deviant attitudes (Rhodes, 1979) and cognitive distortions (De- bowska et al., in press). Consequently, it appears that prisoners are subject to group socialization processes and may develop an iden- tification with the group values. Further, given that social identity must be formed prior to the acquisition of group attitudes (Tajfel & Turner, 1979), a significant correlation between imprisonment and MCSI could be predicted. However, although inconsistent with those past findings, the lack of direct effect of period of incarceration on MCSI in the present study was not entirely

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

low med high

In -g

ro up

T ie

s

Total time spent in prison

Figure 2. The moderating role of interpersonal manipulation in the relationship between time spent in prison and in-group ties subscale of MCSI. The solid line with square markers indicates high ( 1 SD) interpersonal manipulation, the dotted line with triangle markers indicates medium (mean) interpersonal manipulation, and the dashed line with cross markers indicates low (�1 SD) interpersonal manipulation. med � medium.

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

low med high

In -g

ro up

T ie

s

Criminal Friend Index

Figure 3. The moderating role of callous affect in the relationship between criminal friend index and in-group ties subscale of MCSI. The solid line with square markers indicates high ( 1 SD) interpersonal manipulation, the dotted line with triangle markers indicates medium (mean) interpersonal manipulation, and the dashed line with cross markers indicates low (�1 SD) interpersonal manipulation. med � medium.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

436 SHERRETTS, BODUSZEK, AND DEBOWSKA

unexpected. This is because this association was previously found to be moderated by primary psychopathy scores (Boduszek et al., 2016). Along similar lines, the significant effect of incarceration on cognitive distortions pertaining to rape in Debowska et al.’s (in press) study was revealed only after having been controlled for background characteristics (including four psychopathy factors) using propensity score matching technique. Therefore, it seems that the effect of incarceration alone is weak or not sufficient to explain the intensification of both CSI and cognitive distortions in prison contexts.

In support of these findings, the present study found that high interpersonal manipulation scores (1 SD above the mean) com- bined with period of incarceration lead to the increase in MCSI scores. Although Boduszek et al. (2016) reported that the relation- ship between time in prison and CSI is moderated by primary psychopathy scores, the current results provide evidence that this association is affected by interpersonal, rather than affective, psy- chopathic traits and that these two psychopathy facets should be treated as separate dimensions. One possible explanation for the current finding is that psychopaths with strong manipulative ten- dencies are likely to form CSI as an adaptation strategy (Black- burn, 2006). Previous research also found that dispositionally selfish individuals may show concern for in-group members due to the benefits this may entail (De Cremer & Van Vugt, 1999). Therefore, it may be that inmates with high interpersonal manip- ulation scores develop a criminal identity in order to increase their survival chances. Moreover, our results reveal that high interper- sonal manipulation scores affected the relationship between time in prison and in-group ties but not between time in prison and cognitive centrality or in-group affect. Therefore, the depth and genuineness of such a transformation appears questionable. In- deed, Schmid and Jones (1991) proposed that prisoners can con- struct inauthentic temporary identities in order to hide their vul- nerabilities.

It is interesting that the association between time spent in prison and MCSI total score as well as in-group ties score for those low in interpersonal manipulation (1 SD below the mean) was signif- icantly negative, indicating that the direction of the slopes changes between those high and low in interpersonal manipulation. This finding provides further evidence for the claim that individuals are likely to simulate changes in identity through the formation of strong bonds with other offenders and that this can be achieved using interpersonal manipulation skills. According to Cooley’s (1998) conception of the looking-glass self, the way people see themselves is reflective of what others think of them. Therefore, using impression management, they may be able to elicit positive evaluations from others, leading to the maintenance of positive self-esteem (Goffman, 1963, 1990). It thus seems that self- concepts of prisoners lacking in interpersonal manipulation skills may be negatively affected by incarceration; however, this claim remains to be tested.

Criminal friends index was found to have a significant effect on in-group ties, but this was only for those individuals scoring high on callous affect (1 SD above the mean). This is partly supportive of research by Boduszek, Adamson, Shevlin, Mallett, and Hyland (2013), which reported a direct link between antisocial friends and the formation of criminal identity; however, those prior analyses failed to examine the moderating effects of psychopathy factors. As such, our study is the first to demonstrate that as the degree of

involvement with criminal associates before first incarceration increases, so does the strength of in-group ties in prison environ- ment— but only for inmates with elevated callous affect scores. It may be that emotionally shallow individuals create bonds with other criminals because having associates helps in the commission of crimes both outside and inside a prison. Indeed, 40% of inmates were reported as chronic or extreme career offenders while incar- cerated (DeLisi, 2003). Additionally, they were found to be re- sponsible for the occurrence of most serious crimes (e.g., murder, rape, and aggravated assaults), and violent offending was previ- ously associated with increased Factor 1 (affective/interpersonal) psychopathy scores (Serin, 1996). It is interesting that this mod- erating effect of callous affect is not extended to in-group affect (which requires emotional engagement with other group members) or cognitive centrality (for which the criminal identity needs to be perceived as salient). Accordingly, it seems that the involvement with criminal associates and elevated callous affect scores do not influence identity change at a deeper, that is, emotional and cog- nitive, level.

Significant associations between two behavioral dimensions of psychopathy and MCSI total score as well as MCSI factors scores were also found. First, erratic lifestyle had a direct positive effect on in-group ties. This finding is not entirely clear; however, it may be that the need for constant stimulation evidenced by individuals with an erratic lifestyle pattern cannot be satisfied in the con- straints of prison settings. Such inmates, hence, may be drawn to other people in an attempt to find an alternative source of stimu- lation. Further, antisocial behavior facet correlated significantly with MCSI total score and three MCSI factors. It appears that, although speculative at this stage, antisocial tendencies may be a consequence rather than an integral part of a psychopathic person- ality (see Boduszek & Debowska, 2016; Boduszek, Dhingra, Hy- land, & Debowska, 2015; Skeem & Cooke, 2010a, 2010b). That is, individuals with a criminal identity are likely to engage in antiso- cial behavior because actions are largely congruent with the iden- tities assumed (Hewitt, 2003). Given the cross-sectional nature of our study, however, this suggestion must be further explored using longitudinal data.

Finally, we also found a relationship between female gender and MCSI scores as well as in-group ties scores. Although studies on CSI among women are missing, prior research has suggested that women are more likely to form stronger bonds and identification with in-group members because of the greater need to be an accepted and supported member of a group (e.g., B. B. Brown & Lohr, 1987; R. Brown et al., 1986; Kiesner et al., 2002; Newman et al., 2007). This finding has an important practical implication. Specifically, female inmates may benefit from having additional sup- port provisions, such as an increased amount of visits to preserve family bonds and enhance the chances of successful return to society after release from prison.

There are some limitations of the study that need to be acknowl- edged. First, the use of self-report data within a sample of prison- ers whose command of language may be poor could have intro- duced several well-known limitations, such as response bias. This aspect of the study, however, could not be controlled by the researchers. Second, the study did not control for participants’ self-esteem scores. It is therefore unknown whether the intensifi- cation of CSI reported here was due to the need for self- enhancement. Third, the MCSI consists of only eight items that

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

437PSYCHOPATHY, PRISON, AND CRIMINAL SOCIAL IDENTITY

reflect three factors of CSI (with only acceptable Cronbach’s alpha). In order to increase internal reliability of the measure and provide a better coverage of the theoretical construct, the MCSI should be extended (see Hair, Black, Babin, & Anderson, 2010). Additionally, longitudinal studies are needed in order to examine the direction of the relationships among variables. Previous re- search on CSI used male samples only and failed to analyze MCSI dimensions separately; thus, despite the aforementioned limita- tions, the results of the present study expand the current under- standing of CSI.

Overall, findings of the current research provide a substantial contribution to the understanding of the etiology of criminal iden- tity in prison settings. The present results revealed that the rela- tionship between period of incarceration and CSI is moderated by both high and low levels of interpersonal manipulation, and hence a high- and low-risk group was identified. Additionally, criminal friend index was associated with in-group ties for high levels of callous affect. Given that actions are consistent with identities assumed (Hewitt, 2003), studies examining inmates’ self- conceptions can also bring an important insight into their behavior within prison contexts.

References

Barker, J. I. (2005). The self as an internal dialogue: Mead, Blumer, Peirce, and Wiley. American Sociologist, 36, 75– 84. http://dx.doi.org/10.1007/ s12108-005-1010-4

Bate, C., Boduszek, D., Dhingra, K., & Bale, C. (2014). Psychopathy, intelligence and emotional responding in a non-forensic sample: An experimental investigation. Journal of Forensic Psychiatry & Psychol- ogy, 25, 600 – 612. http://dx.doi.org/10.1080/14789949.2014.943798

Blackburn, R. (2006). Other theoretical models of psychopathy. In C. J. Patrick (Ed.), Handbook of psychopathy (pp. 35–57). New York, NY: Guilford Press.

Blumer, H. (1966). Sociological implications of the thought of George Herbert Mead. American Journal of Sociology, 71, 535–544. http://dx .doi.org/10.1086/224171

Boduszek, D., Adamson, G., Shevlin, M., & Hyland, P. (2012a). Devel- opment and validation of a measure of criminal social identity within a sample of Polish recidivistic prisoners. Criminal Behaviour and Mental Health, 22, 315–324. http://dx.doi.org/10.1002/cbm.1827

Boduszek, D., Adamson, G., Shevlin, M., & Hyland, P. (2012b). The role of personality in the relationship between criminal social identity and criminal thinking style within a sample of prisoners with learning difficulties. Journal of Learning Disabilities and Offending Behaviour, 3, 12–23. http://dx.doi.org/10.1108/20420921211236771

Boduszek, D., Adamson, G., Shevlin, M., Hyland, P., & Bourke, A. (2013). The role of criminal social identity in the relationship between criminal friends and criminal thinking style within a sample of recidivistic pris- oners. Journal of Behavior in the Social Environment, 23, 14 –28. http://dx.doi.org/10.1080/10911359.2013.737289

Boduszek, D., Adamson, G., Shevlin, M., Mallett, J., & Hyland, P. (2013). Criminal social identity of recidivistic prisoners: The role of self-esteem, family and criminal friends. Journal of Police and Criminal Psychology, 28, 15–25. http://dx.doi.org/10.1007/s11896-012-9105-7

Boduszek, D., & Debowska, A. (2016). Critical evaluation of psychopathy measurement (PCL–R and SRP-III/SF) and recommendations for future research. Journal of Criminal Justice, 44, 1–12. http://dx.doi.org/10 .1016/j.jcrimjus.2015.11.004

Boduszek, D., Dhingra, K., & Debowska, A. (2016). The moderating role of psychopathic traits in the relationship between period of confinement and criminal social identity in a sample of juvenile prisoners. Journal of

Criminal Justice, 44, 30 –35. http://dx.doi.org/10.1016/j.jcrimjus.2015 .11.005

Boduszek, D., Dhingra, K., & Debowska, A. (in press). The integrated psychosocial model of criminal social identity (IPM-CSI). Deviant Be- havior.

Boduszek, D., Dhingra, K., Hyland, P., & Debowska, A. (2015). A bifac- torial solution to the Psychopathy Checklist: Screening Version in a sample of civil psychiatric patients. Criminal Behaviour and Mental Health. Advance online publication. http://dx.doi.org/10.1002/cbm.1956

Boduszek, D., & Hyland, P. (2011). The theoretical model of criminal social identity: Psycho-social perspective. International Journal of Criminology & Sociological Theory, 4, 604 – 615.

Boduszek, D., & Hyland, P. (2012). Psycho-sociological review of crim- inal thinking style. Journal of Humanistics and Social Sciences, 1, 28 –36.

Boduszek, D., Hyland, P., Bourke, A., Shevlin, M., & Adamson, G. (2013). Assessment of psycho-social factors predicting recidivistic violent of- fences within a sample of male prisoners. Irish Journal of Psychology, 34, 24 –34. http://dx.doi.org/10.1080/03033910.2012.754324

Boduszek, D., Hyland, P., Pedziszczak, J., & Kielkiewicz, K. (2012). Criminal attitudes, recidivistic behaviour, and the mediating role of associations with criminal friends: An empirical investigation within a prison sample of violent offenders. Europe’s Journal of Psychology, 8, 18 –31. http://dx.doi.org/10.5964/ejop.v8i1.296

Boduszek, D., McLaughlin, C., & Hyland, P. (2011). Criminal attitudes of ex-prisoners: The role of personality, anti-social friends and recidivism. Internet Journal of Criminology, 9, 1–10.

Boduszek, D., O’Shea, C., Dhingra, K., & Hyland, P. (2014). Latent class analysis of criminal social identity in a prison sample. Polish Psycho- logical Bulletin, 45, 192–199. http://dx.doi.org/10.2478/ppb-2014-0024

Boduszek, D., Shevlin, M., Adamson, G., & Hyland, P. (2013). Eysenck’s personality model and criminal thinking style within a violent and nonviolent offender sample: Application of propensity score analysis. Deviant Behavior, 34, 483– 493. http://dx.doi.org/10.1080/01639625 .2012.748628

Brown, B. B., & Lohr, M. J. (1987). Peer-group affiliation and adolescent self-esteem: An integration of ego-identity and symbolic-interaction theories. Journal of Personality and Social Psychology, 52, 47–55. http://dx.doi.org/10.1037/0022-3514.52.1.47

Brown, R., Condor, S., Mathews, A., Wade, G., & Williams, J. (1986). Explaining intergroup differentiation in an industrial organization. Jour- nal of Occupational Psychology, 59, 273–286. http://dx.doi.org/10.1111/ j.2044-8325.1986.tb00230.x

Burke, P. J. (2006). Identity change. Social Psychology Quarterly, 69, 81–96. http://dx.doi.org/10.1177/019027250606900106

Cameron, J. (2004). A three factor model of social identity. Self and Identity, 3, 239 –262. http://dx.doi.org/10.1080/13576500444000047

Clemmer, D. (1940). The prison community. Boston, MA: Christopher. Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation

analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. Cooley, C. H. (1998). On self and social organization. Chicago, IL:

University of Chicago Press. Debowska, A., Boduszek, D., Dhingra, K., & DeLisi, M. (in press). The

effect of male incarceration on rape myth acceptance: Application of propensity score matching technique. Deviant Behavior.

Debowska, A., Boduszek, D., Kola, S., & Hyland, P. (2014). A bifactor model of the Polish version of the Hare Self-Report Psychopathy Scale. Personality and Individual Differences, 69, 231–237. http://dx.doi.org/ 10.1016/j.paid.2014.06.001

Debowska, A., Mattison, M. L. A., & Boduszek, D. (2015). Gender differences in the correlates of reactive aggression. Polish Psychological Bulletin, 46, 469 – 476. http://dx.doi.org/10.1515/ppb-2015-0053

Debowska, A., & Zeyrek Rios, E. Y. (2015). The role of psychopathy factors in reactive aggression within a sample of prisoners. Journal of

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

438 SHERRETTS, BODUSZEK, AND DEBOWSKA

Criminal Psychology, 5, 25–33. http://dx.doi.org/10.1108/JCP-10-2014- 0014

Declercq, F., Carter, R., & Neumann, C. S. (2015). Assessing psychopathic traits and criminal behavior in a young adult female community sample using the Self-Report Psychopathy Scale. Journal of Forensic Sciences, 60, 928 –935. http://dx.doi.org/10.1111/1556-4029.12783

De Cremer, D., & Van Vugt, M. (1999). Social identification effects in social dilemmas: A transformation of motives. European Journal of Social Psychology, 29, 871– 893. http://dx.doi.org/10.1002/(SICI)1099- 0992(199911)29:7�871::AID-EJSP962�3.0.CO;2-I

DeLisi, M. (2003). Criminal careers behind bars. Behavioral Sciences & the Law, 21, 653– 669. http://dx.doi.org/10.1002/bsl.531

Dhingra, K., Boduszek, D., Palmer, D., & Shevlin, M. (2015). Psychopathy and self-injurious thoughts and behaviour: Application of latent class analysis. Journal of Mental Health, 24, 4 – 8. http://dx.doi.org/10.3109/ 09638237.2014.910645

Friborg, O., Barlaug, D., Martinussen, M., Rosenvinge, J. H., & Hjemdal, O. (2005). Resilience in relation to personality and intelligence. Inter- national Journal of Methods in Psychiatric Research, 14, 29 – 42. http:// dx.doi.org/10.1002/mpr.15

Friborg, O., Hjemdal, O., Rosenvinge, J. H., & Martinussen, M. (2003). A new rating scale for adult resilience: What are the central protective resources behind healthy adjustment? International Journal of Methods in Psychiatric Research, 12, 65–76. http://dx.doi.org/10.1002/mpr.143

Goffman, E. (1963). Stigma: Notes on the management of spoiled identity. New York, NY: Simon & Schuster.

Goffman, E. (1990). The presentation of self in everyday life. London, United Kingdom: Penguin.

Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Multivariate data analysis: A global perspective. Upper Saddle River, NJ: Pearson Prentice Hall.

Hare, R. D. (1991). The Hare Psychopathy Checklist-Revised. Toronto, Ontario, Canada: Multi-Health Systems.

Hewitt, J. P. (2003). Symbols, objects, and meanings. In L. T. Reynolds & N. J. Herman-Kinney (Eds.), Handbook of symbolic interactionism (pp. 307–326). Lanham, MD: AltaMira Press.

Hogg, M. A., & Reid, S. A. (2006). Social identity, self-categorization, and the communication of group norms. Communication Theory, 16, 7–30. http://dx.doi.org/10.1111/j.1468-2885.2006.00003.x

Holsinger, A. M. (1999). Assessing criminal thinking: Attitudes and ori- entations influence behavior. Corrections Today, 61, 22–25.

Jose, P. E. (2013). ModGraph-I: A programme to compute cell means for the graphical display of moderational analyses: The internet version (Version 3.0) [Computer program]. Wellington, New Zealand: Victoria University of Wellington. Retrieved from http://pavlov.psyc.vuw.ac.nz/ paul-jose/modgraph/

Kiesner, J., Cadinu, M., Poulin, F., & Bucci, M. (2002). Group identifica- tion in early adolescence: Its relation with peer adjustment and its moderator effect on peer influence. Child Development, 73, 196 –208. http://dx.doi.org/10.1111/1467-8624.00400

León-Mayer, E., Folino, J. O., Neumann, C., & Hare, R. D. (2015). The construct of psychopathy in a Chilean prison population. Revista Brasil- eira de Psiquiatria, 37, 191–196. http://dx.doi.org/10.1590/1516-4446- 2014-1540

Levenson, M. R., Kiehl, K. A., & Fitzpatrick, C. M. (1995). Assessing psychopathic attributes in a noninstitutionalized population. Journal of Personality and Social Psychology, 68, 151–158. http://dx.doi.org/10 .1037/0022-3514.68.1.151

Mills, J. F., & Kroner, D. G. (1999). Measures of criminal attitudes and associates. Unpublished user guide.

Newman, B. M., Lohman, B. J., & Newman, P. R. (2007). Peer group membership and a sense of belonging: Their relationship to adolescent behavior problems. Adolescence, 42, 241–263.

Paulhus, D. L., Neumann, C. S., & Hare, R. D. (in press). Manual for the Self-Report Psychopathy Scale. Toronto, Ontario, Canada: Multi-Health Systems.

Rhodes, M. L. (1979). The impact of social anchorage on prisonization. Dissertation Abstracts International, 40, 1694A. (UMI No. 79 –19, 101)

Schmid, T. J., & Jones, R. S. (1991). Suspended identity: Identity trans- formation in maximum security prison. Symbolic Interaction, 14, 415– 432. http://dx.doi.org/10.1525/si.1991.14.4.415

Serin, R. C. (1996). Violent recidivism in criminal psychopaths. Law and Human Behavior, 20, 207–217. http://dx.doi.org/10.1007/BF01499355

Shagufta, S., Boduszek, D., Dhingra, K., & Kola-Palmer, D. (2015). Criminal social identity and suicide ideation among Pakistani young prisoners. International Journal of Prisoner Health, 11, 98 –107. http:// dx.doi.org/10.1108/IJPH-06-2014-0018

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. http://dx.doi.org/10.1177/0093854897024001004

Simourd, D. J., & Van De Ven, J. (1999). Assessment of criminal attitudes: Criterion-related validity of the Criminal Sentiments Scale–Modified and Pride in Delinquency scale. Criminal Justice and Behavior, 26, 90 –106. http://dx.doi.org/10.1177/0093854899026001005

Skeem, J. L., & Cooke, D. J. (2010a). Is criminal behavior a central component of psychopathy? Conceptual directions for resolving the debate. Psychological Assessment, 22, 433– 445. http://dx.doi.org/10 .1037/a0008512

Skeem, J. L., & Cooke, D. J. (2010b). One measure does not a construct make: Directions toward reinvigorating psychopathy research—Reply to Hare and Neumann (2010). Psychological Assessment, 22, 455– 459. http://dx.doi.org/10.1037/a0014862

Stryker, S., & Burke, P. J. (2000). The past, present, and future of an identity theory. Social Psychology Quarterly, 63, 284 –297. http://dx.doi .org/10.2307/2695840

Swann, W. B., Jr. (1987). Identity negotiation: Where two roads meet. Journal of Personality and Social Psychology, 53, 1038 –1051. http:// dx.doi.org/10.1037/0022-3514.53.6.1038

Tajfel, H., & Turner, J. (1979). An integrative theory of intergroup conflict. In W. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33– 47). Monterey, CA: Brooks/Cole.

Turner, J. C. (1982). Towards a cognitive redefinition of the social group. In H. Tajfel (Ed.), Social identity and intergroup relations (pp. 15– 40). Cambridge, United Kingdom: Cambridge University Press.

Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social groups: A self-categorization theory. Oxford, United Kingdom: Blackwell.

Vryan, K. D., Adler, P. A., & Adler, P. (2003). Identity. In L. T. Reynolds & N. J. Herman-Kinney (Eds.), Handbook of symbolic interactionism (pp. 367–390). Lanham, MD: AltaMira Press.

Weigert, A. J., & Gecas, V. (2003). Self. In L. T. Reynolds & N. J. Herman-Kinney (Eds.), Handbook of symbolic interactionism (pp. 267– 288). Lanham, MD: AltaMira Press.

Received June 28, 2015 Revision received February 16, 2016

Accepted February 20, 2016 �

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

439PSYCHOPATHY, PRISON, AND CRIMINAL SOCIAL IDENTITY