Assessing the Relationship Between Religiosity and Recidivism Among Adult Probationers in Pakistan

Abstract

Although empirical evidence supports a relationship between religiosity and criminal behavior, debate continues about the theoretical mechanisms by which they are related. Moreover, the topic has been largely ignored by practicing clinicians and correctional workers. The Muslim Religiosity-Personality Inventory: Abridged was administered to low-risk Pakistani probationers and factor analyzed, after which probationers’ recidivism was monitored. Five oblique factors were obtained, three of which were correlated with recidivism (Religious Practice, Religious–Moral Values, and Fundamental Religious Beliefs), as was the full scale, while two were not (Importance of Religion and Rejection of Nonbeliever). In a logistic regression, Religious–Moral Values and Religious Practices contributed to the prediction of probationer recidivism. However, when demographic characteristics were introduced, education and marital status replaced Religious Practices. This study supports the religiosity–crime link in a non-Western, Muslim culture. Implications for assessing religiosity and for practitioners in the justice system are discussed.

Keywords

religiosity probation Muslim Religiosity-Personality Inventory Muslim offenders offender risk assessment

Given what might be an intuitive sense about the nature of religion, criminal behavior, and the potential for their interrelation (Heaton, 2006), one might be surprised by the relatively limited amount of research and correctional practice devoted to this topic (Johnson, 2008). For example, religion is not even mentioned in a number of classic criminology texts (Cullen, 2010), and the concept of religiosity is almost nonexistent in both offender assessment research and clinical practice. Nonetheless, some theoretical and empirical work has generated various perspectives on the crime–religion connection in the domains of sociology, criminology, forensic psychology, and behavioral genetics (Baier & Wright, 2001; Beaver, Gibson, Jennings, & Ward, 2009; Benda & Toombs, 2002; Benda, Toombs, & Peacock, 2003; Evans, Cullen, Dunaway, & Burton, 1995; Gottfredson & Hirschi, 1990; Johnson, Jang, Larson, & De Li, 2001; Sinha, Cnaan, & Gelles, 2007).

Religiosity and Criminal Behavior

The idea that religion and criminal behavior are somehow related has an intuitive appeal (Heaton, 2006), and a number of sociological, psychological, and criminological theories have offered explanations of the complex relationship between religion and crime. They include social control theory (Johnson, Larson, De Li, & Jang, 2000), differential association theory (Wright, Caspi, Moffitt, & Silva, 1999), and rational choice theory (Grasmick, Bursik, & Cochran, 1991). Moreover, in their “psychology of criminal conduct” (PCC), which draws upon social learning, psychological, and social-cognition theories to explain human variation in criminal behavior, Bonta and Andrews (2017) include certain leisure activities, such as “participation on a committee for a religious organization” (Andrews, Bonta, & Wormith, 2004, p. 17), as strengths or protective factors against the likelihood of reoffending. On the contrary, the theoretical perspective of Hirschi and Stark (1969) and later work by their colleagues (Burkett & White, 1974) have been instrumental in propagating the notion that religion has little or no direct association with criminality. However, these researchers have accepted some indirect link between religion/faith and crime by suggesting that religion strengthens an individual’s bonds with social institutions and society that ultimately affect deviant behavior (see Baier & Wright, 2001; Evans et al., 1995; Gottfredson & Hirschi, 1990).

Since 1969, an impressive quantity of literature, including individual studies, meta-analytic investigations, and research reviews, have produced findings supporting the notion that religiosity and its various dimensions (e.g., church attendance, perception of God or religious salience, participation in religious activities) often have an inverse relationship with antisocial behavior (e.g., delinquency, substance use, violence, or parole/probation violations) among various samples, including inmates (Benda, 1995; Benda et al., 2003), community offenders (Bhutta & Wormith, 2016; Jensen & Gibbons, 2002; Tittle & Welch, 1983), youth (Johnson et al., 2001; Johnson et al., 2000; Kerley, Copes, Tewksbury, & Dabney, 2010; Sinha et al., 2007), and adults (Evans et al., 1995). In addition, there are several research studies and reviews investigating the link between religiosity and delinquency or violence among youth and adolescent populations (Baier & Wright, 2001; Salas-Wright, Vaughn, Hodge, & Perron, 2012; Sinha et al., 2007; Yeung, Chan, & Lee, 2009). Moreover, there is growing support that religion may be a useful tool to help prevent high-risk urban youth from engaging in delinquent behavior (Johnson et al., 2000). While most of the research has examined youth populations, fewer studies have explored the relationship between the various dimensions of religion and adult criminality.

Among adult offenders, Evans et al. (1995) assessed the impact of religiosity on adult criminal behavior, using a 43-item measure on a sample of White males from an urban area in the United States. By assessing various dimensions of religion, including “hellfire beliefs,” religious attitudes, and participation in religious activities, they found that participation in religious activities was the only dimension that had a direct effect on reducing criminal behavior. They also reported that participation in religious activities was negatively linked to a broad range of criminal acts (Evans et al., 1995). Thus, they supported the conclusions of previous studies investigating the relationship between religious services and delinquency (Higgins & Albrecht, 1977; Tittle & Welch, 1983). Moreover, research by Benda and colleagues, on a sample of Southern boot camp inmates in the United States, supported a religiosity–crime link (Benda, 1995; Benda et al., 2003). They demonstrated that not only lower levels of religiosity predicted increased violent offending (Benda & Toombs, 2000) but also level of religiosity differentiated nonrecidivists from offenders responsible for various kinds of recidivism (i.e., those with felonies, technical parole violators, or drug-related parole violations; Benda et al., 2003). Jensen and Gibbons (2002) used a sample of felonious/serious prisoners, serving between 5 and 25 years, to conduct a qualitative study on the effects of religion on recidivism (i.e., parole violations). Their study supported the link between religion and criminal recidivism in that religion played a significant role in the reintegration process among those offenders who self-reported being religious. These offenders were considerably less likely to commit either parole violations or new offences (Jensen & Gibbons, 2002).

Much of this research is captured in a comprehensive meta-analysis by Baier and Wright (2001) that investigated the relationship between religion and crime among samples of incarcerated (mostly felonious) offenders. A variety of outcome measures of offending behavior (i.e., violence, drug use and sale, probation violation, weapon carrying) were examined. Based on 60 studies on the religion–crime relationship conducted between 1969 and 1998, they found a moderate, negative effect size (r = −.12) of religion on antisocial and criminal behavior. Although they discussed sample size, offence type, race, geographic region, and criminal activity (the dependent variable) at considerable length, little attention was given to the measures of religion used in these studies. More recently, Johnson and Jang (2010) reported that religiosity served as a significant deterrent against antisocial behavior in 90% of the 270 they reviewed.

Yet gaps remain in the research pertaining to the relationship between religiosity and criminal or antisocial behavior in theoretical explanations for it. With respect to theoretical explanations of the religiosity–crime link, most have relied on popular criminological theories and, to some extent, social psychological theories (Watts, 2018). These include self-control theory (Gottfredson & Hirschi, 1990), social bond theory (Hirschi, 1969), and social learning theory (Bandura, 1977), with examples of their application found in research by Baier and Wright (2001), Cretacci (2003), and Adamczyk (2012), respectively. With Hirschi’s (2004) efforts to encompass various kinds of inhibitions (internal and external) on behavior, we have a more integrated perspective of self-control, but still with various components, including attachment, commitments, involvement, and beliefs, from which we may consider the religiosity–crime relationship.

With respect to the extant research, less attention has been paid to low-risk/nonviolent offenders (specifically adults), most of the research has used single- or two-item measures of religiosity, and a Judeo-Christian conceptual framework of religiosity has been commonly used when examining the association between religion and criminal behavior. Given the complexity of the religiosity construct and the paucity of religiosity instruments developed upon Muslim samples, the Judeo-Christian religiosity scales have been challenged in terms of appropriateness for measuring religiosity among culturally diverse Muslim populations (Abu Raiya, Pargament, Mahoney, & Stein, 2008; Amer & Bagasra, 2013; Krauss, Hamzah, & Idris, 2007; Krauss et al., 2006). These Judeo-Christian religiosity constructs may not be appropriate when examining the religiosity level of non-Western samples of adult offenders (probationers) in a developing country. As an examination of religiosity will necessarily depend on the measurement of religiosity, attention to the psychometric properties of non-Christian measures, such as the Muslim Religiosity-Personality Inventory (MRPI; Krauss et al., 2006), is required if the religiosity–crime relationship is to be studied in non-Western cultures. This includes Pakistan, which is a highly religious country where more than 97% of the population is Muslim and there is a unique, socio-religious culture (Hassan, 2007).

Probation in Pakistan

The Probation of Offenders Ordinance (Pakistan, 1960) was introduced in all provinces of Pakistan in 1960 to provide a legal framework allowing the judiciary/courts to place or release eligible offenders (e.g., first-time and nonviolent offenders) on probation for no more than 3 years under certain conditions. The probation law in Pakistan is intended to rehabilitate and reintegrate “petty offenders” into the community (Rules 10/1961; Pakistan, 1960) by providing them the opportunity to make themselves “honest, industrious and law-abiding” individuals of society under the supervision, assistance, and cooperation of probation officers (POs). POs are employed in all 36 districts in the Punjab province under the administrative control of the Reclamation and Probation (R&P) Department (Rules 3-6, 1961; Pakistan, 1960). As is the case in most Western judicial settings, the court seeks the assistance of the POs following the conviction of petty offenders when making decisions to release offenders on probation. POs prepare and submit a Social Investigation Report (SIR) about the offender that includes the nature of the offence, the antecedents, the character of the offender, their home environment, and any other important matters related to the commission of the offence (Bhutta, 2010; Hamid-uz-Zafar, 1961). In principle, this report is expected to provide comprehensive evidence for the court to pass a probation order for the eligible offender. It is also notable that no risk/need assessment is currently used to determine the level of risk and needs of offenders objectively in the Punjab province. This is unlike most Western jurisdictions that use such instruments to aid in writing the SIR, to make judicial decisions, to conduct offender classifications, to carry out case management, and to offer correctional treatment to offenders after their release on probation or parole (Andrews, Bonta, & Wormith, 2006).

In sum, the role of the PO is pivotal in the probation process. It begins before the probationer’s release and continues through the formulation of a supervision and guidance strategy during the rehabilitative process. It is also the duty of POs to inform the courts of any instances of probationers’ failure to observe the conditions of the probation order (e.g., probationers’ attendance in the POs’ office is obligatory on at least a month basis). The Probation of Offenders Ordinance 1960, Section 7 (Pakistan, 1960) also empowers the courts to revoke a probation order by issuing a new order to rearrest probationers who have failed to observe any of the probation conditions.

The Present Study

The current study examined the relationship between religiosity or religious dimensions (factors) and recidivism on a sample of adult Pakistani probationers. More specifically, it assessed whether various dimensions of religiosity, such as religious beliefs, values, practices/rituals, and participation in religious congregations, were related to probationers’ recidivism. The study was guided by the following research questions:

Research Question 1: What are the psychometric properties, including factor structure, of a brief religiosity measure (MRPI; Krauss et al., 2006), which was adapted for this population?

Research Question 2: What is the nature of the relationship between religiosity and recidivism in a Muslim culture?

Research Question 3: Do the religiosity factors have different degrees of association with recidivism?

Research Question 4: Do sociodemographic variables, in combination with offence-type variables, affect the association between religiosity factors and recidivism?

The study was exploratory in nature, both theoretically and psychometrically, with no specific hypotheses aside from the anticipated religiosity–recidivism relationship. However, post hoc theoretically relevant commentary is offered, particularly with respect to the derived religiosity factors and their relationship with recidivism.

Method

Participants

The sample consisted of 506 adult probationers (18 years of age and older), released on probation during the year 2010. As assessments are mandated by law, they represent virtually all probationers in the Lahore Division, which consists of four districts in the Punjab province of Pakistan. This division was selected in consultation with officials from the Directorate of R&P Department as being representative of Punjab. The probationers consisted of subgroups defined by religion/faith (Muslim [n = 475, 93.9%] and Christian [n = 31, 6.1%]), gender (male [n = 470, 92.9%] and female [n = 36, 7.1%]), and geographical location (urban [n = 288, 56.9%] and rural [n = 218, 43.1%]). Probationers who resided in cities, municipalities, and town communities in Punjab were categorized as urban, while all others were categorized as rural. With respect to the offence type, 51.4% of the probationers were involved in drug-related offences, followed by theft (26.9%), carrying a weapon (11.5%), and miscellaneous offences (10.3%). The latter three groups were merged into a nondrug group (n = 246, 48.6%) for further statistical comparisons. The participants’ mean age was 32.47 years (SD = 11.06), with a range of 18 to 65 years. There was no difference in age for males and females. However, urban probationers were significantly older than rural probationers, t(504) = 2.34, p < .02. With regard to education, almost one half of the total sample (n = 226, 44.7%) had no formal education and was illiterate. The mean monthly income was reported to be 6,665 PKR (SD = 5,077) or approximately US$67. As a group, the offenders may be described as low risk, in part because they were all first-time offenders, as required by Pakistani law to qualify for probation, and in part because their mean score on a commonly used risk/need assessment tool, the Level of Service/Case Management Inventory (LS/CMI; Andrews et al., 2004), which was adapted for Pakistani culture, fell in the “low risk” range (M = 8.58, SD = 4.46; Bhutta & Wormith, 2016).

Measures

Muslim Religiosity-Personality Inventory: Abridged (MRPI-A)

Items from the Religious Personality subscale (Krauss et al., 2007) of the 102-item MRPI (Krauss et al., 2006) were adapted or modified to create an abridged form (14 items) of the MRPI and investigated for its psychometric properties. This was done as part of the current investigation of general religiosity and its various dimensions in a sample of adult probationers in Pakistan. A brief version of the religiosity measure was sought for operational and practical reasons because time with offenders was limited and many had limited or no education.

The original version of the MRPI (Krauss & Hamzah, 2009; Krauss et al., 2006) was developed in Malaysia for the Muslim youth population by using the Muslim Religious Personality Measurement model (Krauss, Hamzah, Juhari, & Hamid, 2005). Unlike scales developed by other researchers (Spilka, Hood, Hunsberger, & Gorsuch, 2003), Krauss and his colleagues (2006) focused completely on the Islamic perspective of religiosity during the scale development process (Krauss et al., 2005). The original version of MRPI is divided into two subscales, the Islamic Worldview subscale and the Religious Personality subscale. The Islamic Worldview subscale assesses

one’s level of agreement with the statements relating to the Islamic pillars of faith (arkan al-Iman, i.e., belief in: God, Angels, Messengers and Prophets of God, Books of Revelation, The Day of Judgment, and the Divine Decree), which represent the foundation of the Islamic creed (aqidah). (Krauss, Hamzah, Juhari, & Hamid, 2005, p. 177)

The Religious Personality subscale involves behaviors, motivations, attitudes, and emotions that are intended to assess personal (intrinsic) manifestations of the Islamic teachings and commandments. This subscale is centered on the formal and ritual worship that reflects one’s direct relationship with God and the daily mu’amalat, or the religiously guided behaviors toward one’s family, fellow human beings, and the rest of creation (i.e., animals and the natural environment) and is known as the general worship or general “ibadat” (Krauss et al., 2006). Interestingly, a number of studies on this measure of religiosity have modified this subscale for use with numerous samples of different age groups, geographic regions, and diverse faiths/religions (Annalakshmi & Abeer, 2012; Krauss et al., 2012). The adapted version of Religious Personality subscale contains 99 items (three items were removed from original 102 items; see details in Krauss et al., 2007).

The item selection strategy for the development of a religiosity scale in the current investigation was to choose items from the Religious Personality subscale that were used repeatedly in prior empirical studies on the religion–crime relationship (e.g., attendance at church or mosque and participation in religious activities). Items were translated into Urdu by the first author and some items were then modified to suit Pakistani language and culture. Specifically, 11 of the 14 items (appendix) in the new religiosity scale were taken either directly from the Religious Personality subscale (nine items) or highly related items were merged into a single item (two items, that is, Items 6 and 12). To examine attitudinal aspects of religiosity (e.g., religious beliefs, importance of religion in life) among participants, three items were adapted from different empirical studies on the religion–crime context (i.e., Items 1, 5, and 14; Johnson et al., 2000; Sinha et al., 2007). Although specific religious concepts that were examined in the scale (e.g., the Prophet, life hereafter, prayer, rights of others, religious gathering, repentance, and forgiveness) are found in Islam, they are also common to other religions including Christianity. Moreover, care was taken in the wording of items to be inclusive of other religions.

In sum, the brief religiosity scale used in this research was termed the MRPI-A scale. It consists of 14 items, with 13 items on a 5-point scale, which were typically coded as “not at all” (1), “minimally” (2), “moderately” (3), “quite a bit” (4), and “very much” (5), as per the original scaling of the MRPI. Due to the nature of the question, one item (Item 2) was coded on a 3-point scale: “no” (1), “perhaps” (2), and “yes” (3). Therefore, the minimum possible score of respondents on this revised scale was 14 and the maximum possible score was 68. Higher scores on the MRPI-A indicate a higher degree of religiosity.

Criminal recidivism

Recidivism was the outcome variable used in this study and was defined as the violation/or revocation or failure on conditions of the probation order as determined by the Probation of Offenders Ordinance (Pakistan, 1960). It included any conviction for a new offence. Data were provided to the researcher in a yes/no format from R&P Department of Punjab after completion of the follow-up period for the entire sample. The follow-up period spanned from 10 to 11 months from the initial offender assessment (i.e., January and February 2011) to the recidivism investigation date (i.e., December 2011).

Procedure

A quantitative survey was conducted to collect data from the four districts of the Lahore division of the Punjab province of Pakistan.¹ To gather information for the research, formal approval was obtained from the R&P Department of Punjab. As part of a larger investigation (Bhutta, 2013), face-to-face interviews of adult probationers were conducted in the offices of the POs in each district by a team of researchers from the Punjab University, Lahore, under the supervision and monitoring of the first author. The MRPI-A was administered to all participants in a paper-and-pencil format and was read to participants who were illiterate. A risk/need assessment inventory, the LS/CMI (Andrews et al., 2004), was also administered to the probationers, the results of which are reported elsewhere (Bhutta & Wormith, 2016). Interviews typically required 30 to 35 min to complete. Socioeconomic, demographic, and legal characteristics of probationers were obtained from official records and files of the participants which were maintained in POs’ offices in each district. In keeping with cultural norms, a female researcher conducted the interviews of female probationers in the office of a female PO. Although a precise measure of the participation rate was not obtained, consistent with Pakistani culture and its tradition of compliance, it was very high (i.e., close to 100%).

This was a prospective investigation, beginning with intake assessments and data collection of offenders early in their term of probation (January and February 2011). A follow-up investigation was conducted at the end of 2011 to determine whether they had recidivated.

Analytic Strategy

Frequency tables and chi-square analyses were conducted to produce descriptive statistics about the entire sample and for various subgroups. A series of t tests, chi-squares, and analyses of variance (ANOVAs) were conducted to determine the variation in religiosity between demographically defined subgroups (i.e., gender, marital status, geographic location, and offence type) and recidivists. Cronbach’s (1951) alphas were calculated to examine the reliability (internal consistency) of items in the religiosity scale and the religious factors. A principal component analysis, using the promax (oblique) rotation method, was conducted to determine the factor structure of the MRPI-A scale. Factor-based subscales were generated based on item factor loadings greater than .40 and labeled in accordance with the authors’ interpretation of common features in the factor items. Pearson’s correlations were also calculated to investigate relationships between overall religiosity, religious factors, background variables, and recidivism. Finally, stepwise logistic regression analyses were conducted to determine the contribution of religious factors, with and without sociodemographic factors, to the prediction of recidivism in the current sample of adult Pakistani probationers. The data analysis was conducted using SPSS (Version 21).

Results

Descriptive Statistics of the MRPI-A Scale

Factor analysis

A principal component analysis using an oblique rotation (Promax, with κ = 4) method with Kaiser normalization was conducted to determine the factor structure of the MRPI-A scale. A five-factor solution, with a factor loading cut-off greater than .40 and Eigen values greater than 1.00, was considered an optimal fit for the MRPI-A scale, as it accounted for 60% of the original variance (Table 1).

Table 1.

Principal Component Analysis Factor Loadings of the MRPI: Abridged Scale (14 Items) and Explained Variance of Factors.

MRPI: abridged items	Factor 1: Religious Practice	Factor 2: Religious–Moral Values	Factor 3: Importance of Religion	Factor 4: Fundamental Religious Beliefs	Factor 5: Rejection of Nonbeliever
6. Religious knowledge	.881	−.252	.018	.041	−.072
8. Pray regularly	.771	.002	.065	−.080	.016
3. Follow beliefs	.599	.178	−.152	.015	.265
4. Follow prophet	.525	.255	−.159	.272	.019
10. Participate in ceremony	.511	−.017	.400	.033	−.115
13. Belief in tolerance	−.250	.893	.046	−.089	−.152
12. Care for rights	.143	.717	.055	.073	−.126
11. Avoid falsehood	.127	.466	.150	−.166	.222
5. Importance of religion	−.106	−.018	.694	.332	.037
14. Repent after wrong	.221	.034	.643	−.320	−.187
9. Prayer prevents	−.121	.229	.619	.084	.266
2. Belief accountable	.032	.082	−.087	.776	−.182
1. Belief in God	.007	−.197	.281	.734	.037
7. Dislike nonbelievers	.014	−.155	.029	−.121	.942
Percentage of variance	(24.68)	(10.39)	(9.22)	(8.23)	(7.57)

Note. Factor loadings greater than .400 presented in bold. Rotation method: Promax with Kaiser Normalization. Rotation converged in eight iterations. See the appendix for complete items.

Meticulous effort was made to label each factor in keeping with the theoretical underpinnings of the items that loaded on each respective factor. In sum, 14 items of the MRPI-A scale produced five religiosity factors labeled as follows: Religious Practices, Religious–Moral Values, Importance of Religion, Fundamental Religious Beliefs, and Rejection of Nonbeliever (Table 1). The component correlation matrix indicated that the first two religious factors were intercorrelated relatively higher (.46) than the other religious factors. However, the correlation was not high enough to be grouped into a single factor, and it was agreed by the research team that Religious Practice and Religious–Moral Values may constitute related, but importantly different aspects of religiosity.

Thirteen items loaded on one factor only as determined by factor loadings greater than .40, while one item (Item 10; Participate in Religious Activities) also loaded on a second factor (Importance of Religion) at .40. The 14 items were then grouped in accordance with their factor loadings greater than .40 and used as subscales in subsequent statistical analyses. As correlated factors were allowed, the first two factors, Religious Practices and Religious–Moral Values, were highly related (r = .41), with the former also correlating with the remaining factors (.272 to .142). These two factors also accounted for 35% of the 60% variance accounted for by the model. Religious–Moral Values, Importance of Religion, and Fundamental Religious Beliefs were modestly correlated with each other (.157 to .261), while factor score correlations with Rejection of Nonbeliever were low (.049 to .142; see Table 2). When the factor analysis was repeated for Muslim probationers only (N = 475), the same factors and pattern matrix were produced and displayed similar weightings.

Table 2.

Correlations Between Legal-Demographic Measures, Religiosity Scores, and Recidivism.

Measure	Age	Educat.	Inc.	Drug	Theft	Weap.	Male	Marry	Urban	F1	F2	F3	F4	F5	Relig.
Educat.	−.080					.
Income	.152**	.161**
Drug	.013	−.190**	−.069
Theft	.065	.068	.081	−.623**
Weapon	−.166**	.165**	.046	−.370**	−.218**
Male	.056	.129**	.287**	−.008	−.040	.075
Married	.456**	−.084	.126**	.002	.078	−.148**	−.038
Urban	.104*	.046	.046	.224**	−.121**	−.176**	−.023	.029
F1	−.028	.276**	.100*	−.221**	.037	.253**	.032	−.029	−.012
F2	−.029	.064	.117**	−.125**	.044	.100*	.046	.013	−.020	.420**
F3	.058	.154**	.022	−.065	−.021	.027	.016	.007	.069	.272**	.261**
F4	−.022	.025	.083	.060	−.117**	−.007	.062	.064	−.016	.212**	.157**	.222**
F5	−.003	−.043	−.021	.034	−.067	−.037	−.010	.067	−.015	.142**	.092*	.090*	.049
Relig.	−.014	.220**	.103*	−.178**	.005	.183**	.04	.010	−.001	.866**	.664**	.543**	.319**	.398**
Recid.	.036	−.131**	−.080	.090*	−.029	−.045	.004	−.077	.056	−.141**	−.144**	−.065	−.115**	.021	−.150**

Note. Educ. = education; Inc. = income; Weap. = weapon; F1 = Religious Practice (five items); F2 = Religious–Moral Values (three items); F3 = Importance Of Religion (three items); F4 = Fundamental Religious Beliefs (two items); F5 = Rejection of Nonbeliever (one item); Relig. = religiosity; Recid. = recidivism.

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

There was considerable variability in mean scores for the religiosity factors as seen in Table 3. This was partly due to their varying number of items, and also due to differences in the mean item scores for the five factors. Mean item scores for factors were significantly different from each other, except Religious–Moral values and Fundamental Religious Beliefs. Items pertaining to the Importance of Religion factor had the highest mean scores, while the single-item factor, Rejection of Nonbelievers, had the lowest mean score.

Table 3.

Descriptive Statistics and Cronbach’s Alpha (α) Coefficients for Total Religiosity and Religious Factors Based on Items With Factor Loadings Greater Than 40.

Religiosity scales (items)	ScaleM	ScaleSD	ItemM	ItemSD	α
Religiosity (14 items)	50.58	6.44	3.61	0.46	.725
1. Religious Practices (five items)	15.02	3.98	3.00	0.80	. 748
2. Religious–Moral Values (three items)	11.77	1.95	3.92^a	0.65	.537
3. Importance of Religion (three items)	13.37	1.66	4.46	0.55	. 481
4. Fundamental Religious Beliefs (two items^b)	7.86	0.46	3.93^a	0.62	.411
5. Rejection of Nonbelievers (one item)	2.55	1.65	2.55	1.65	nil

All paired comparisons of factor item means are different from each other at p < .001, except where there is a common superscript.

Includes one item on a 1 to 3 scale.

Reliability

Cronbach’s (1951) alpha was used to examine the internal consistency of the MRPI-A scale (14 items) and the five religiosity factors. The internal consistency of factors, based only on items loading on factors by more than .40, was compared with the alphas of the complete scale (Table 3). For the overall religiosity scale, the analysis yielded a moderate level of reliability (α = .725) for the whole sample. However, Cronbach’s (1951) alpha values varied considerably for the five religiosity factors. Factor 1, consisting of five items, had a slightly higher alpha (α = .748) than the complete 14-item scale. Factors 2, 3, and 4, with only two or three items each, generated low alphas (α = .411 to .537). This is not unexpected given their limited number of items and raises questions about these factors functioning effectively as subscales of the MRPI-A scale. The Guttman split-half correlation coefficient based on all 14 items in the full scale was .760.

Descriptive statistics and group comparison on the MRPI-A scale

The mean score on the MRPI-A scale for the entire sample was 50.56 (SD = 6.44; Table 3). Mean item scores differed significantly across the five factors, with Importance of Religion items being most highly rated, followed by Fundamental Religious Beliefs items (despite its inclusion of an item on a 3-point scale), Religious–Moral Values items, Religious Practices items, and finally the Rejection of Nonbeliever items.

There was no significant difference between Muslim (M = 50.61, SD = 6.49) and Christian (M = 49.83, SD = 5.63) groups on the MRPI-A total score, t(504) = .646, p = .519. Muslims (M = 2.58, SD = 1.65) tended to score more highly than Christians (M = 2.03, SD = 1.52) on one of the 14 items, specifically Item 7, dislike for nonbelievers, t(504) = −1.806, p = .071. The mean MRPI-A score was not statistically different across gender (male [M = 50.63, SD = 6.49] and female [M = 49.64, SD = 5.84]), t(504) = 0.893, p = .519, geographical location (urban [M = 50.56, SD = 6.49] and rural [M = 50.57, SD = 5.84]), t(504) = −0.017, p = .987, and marital status (married [M = 50.61, SD = 6.03] and unmarried [M = 50.49, SD = 6.97]), t(504) = 0.216, p = .829. However, drug-related offenders had significantly lower religiosity scores than the pooled group of other nondrug offenders (drug-related offenders [M = 49.45, SD = 5.81] and nondrug-related offenders [M = 51.74, SD = 6.86]), t(504) = −4.06, p < .001.

Recidivism

Only 44 (8.7%) probationers in the sample of 506 probationers recidivated during the follow-up period. Muslim (8.8 %) and Christian (6.5 %) probationers did not differ in their recidivism rate, t(504) = 0.46, p = .65. These recidivism rates are considered low compared with similar research of adult probationers in the United States (e.g., Maruschak & Bonczar, 2014) and Canada (e.g., Girard & Wormith, 2004). Two likely reasons for the low recidivism rate were the relatively short follow-up period and the stringent requirements for probation in Pakistan (i.e., first offenders only), which would render probationers to be low risk from the outset.

Recidivists were also compared with nonrecidivists with respect to age, education, and income as potential mediators of a religiosity–crime link. The mean education level of nonrecidivists (M = 4.11, SD = 4.164) was significantly higher than that of recidivists (M = 2.20, SD = 3.070), t(506) = −2.960, p = .003. Income was marginally significant in differentiating recidivists (M = 5,352.27, SD = 4,255.75) from nonrecidivists (M = 6,790.26, SD = 5,135.87), t(506) = −1.799, p = .07. There was no difference between the recidivists (M = 32.35, SD = 11.131) and the nonrecidivists (M = 33.77, SD = 10.320) with respect to age, t(506) = 0.816, p =.415. However, there was a significant difference between recidivists (M = 47.43, SD = 6.88) and nonrecidivists (M = 50.86, SD = 6.32) on the MRPI-A score, t(504) = −3.410, p = .001, indicating that the MRPI-A scale demonstrated a capacity to differentiate recidivists from nonrecidivists among adult probationers in Pakistan.

Point-biserial correlations

There was a small, but significant inverse relationship between the overall religiosity score and recidivism (r = −.15, p = .001), indicating that higher religiosity scores were related to a lower probability of reoffending (Table 3). When the relationship between the religiosity factor scores and recidivism was analyzed separately, three religiosity factors (e.g., Religious Practices, Religious–Moral Values, and Fundamental Beliefs) produced statistically significant inverse correlations (−.14, −.14, and −.12, all ps < .01) with reoffending. Among the sociodemographic and offence-type variables, only education (r = −.13, p = .01) and offence group (i.e., drug-related; r = .09, p = .05) produced significant point-biserial correlations with recidivism for the entire sample.

Religiosity factors alone and with demographic variables

Logistic regression analyses were conducted to examine the association between the five religiosity factors and recidivism. Using the entire sample, the five religiosity factor scores were entered in a stepwise (forward conditional) logistical regression in an effort to explain the variation in the outcome measure, recidivism (Table 4, Model 1). Independent variables were entered sequentially if they met a .05 level of significance and removed if they failed to meet a .10 level of significance. Two religiosity factors, namely, Religious–Moral Values (odds ratio [OR] = .905, 95% confidence interval [CI] = [.824, .993]) and Religious Practices (OR = .847, 95% CI = [.719, .997]), were significantly related to recidivism. The negative sign of the Beta coefficients indicated that increased scores of these two religiosity factors were simultaneously associated with reduced probability of reoffending among the current sample. These findings were reminiscent of the correlation results for the five religiosity factors, with the exception that one religiosity factor (i.e., Fundamental Religious Beliefs), which produced a significant correlation (r = −.115, p = .009; Table 2) with recidivism, was not included in the regression model, indicating that its relationship with recidivism was shared with Religious Practices and Religious Values. The remaining two religiosity factors (i.e., Importance of Religion and Rejection of Nonbelievers) were not related to reoffending in either the correlation or the regression analyses.

Table 4.

Forward Stepwise (Conditional) Logistic Regression of Religiosity Factors and Demographic Measures on Probationer Recidivism.

Religious and religious/demographic factors		B	SE	Wald	df	p	Exp (B)(lower to upper)
Model 1
Step 1	Religious–Moral Values	−0.242	.076	10.075	1	.002	0.787 [0.678, 0.912]
	Constant	0.380	.846	0.202	1	.653	1.462
Step 2	Religious–Moral Values	−0.100	.048	4.425	1	.035	0.905 [0.824, 0.993]
	Religious Practices	−0.167	.084	3.970	1	.046	0.847 [0.719, 0.997]
	Constant	0.964	.906	1.132	1	.287	2.623
Model 2
Block 1
Step 1	Education	−0.132	.046	8.122	1	.004	0.876 [0.800, 0.960]
	Constant	−1.943	.192	102.176	1	.000	0.143
Step 2	Education	−0.139	.047	8.862	1	.003	0.871 [0.795, 0.954]
	Marital status	−0.629	.321	3.843	1	.050	0.533 [0.284, 1.000]
	Constant	−1.579	.255	38.251	1	.000	0.206
Block 2
Step 1	Education	−0.119	.048	6.152	1	.013	0.888 [0.808, 0.975]
	Marital status	−0.549	.325	2.850	1	.091	0.578 [0.306, 1.092]
	MRPI: Abridged total	−0.987	.365	7.303	1	.007	0.373 [0.182, 0.763]
	Constant	1.779	1.245	2.042	1	.153	5.924
Model 3^a
Block 2
Step 1	Education	−0.135	.048	8.063	1	.005	0.873 [0.796, 0.959]
	Marital status	−0.611	.324	3.552	1	.059	0.543 [0.288, 1.025]
	Religious–Moral Values	−0.234	.078	9.070	1	.003	0.791 [0.679, 0.921]
	Constant	1.064	.896	1.412	1	235	2.899

Note. Regressions coefficients are as follows—Model 1, Step 1: –2 log likelihood = 289.182, χ²(1) = 9.803, p = .002, Cox and Snell R² = .019, Nagelkerke R² = .043; Model 1, Step 2: –2 log likelihood = 284.555), χ²(2) = 14.430, p = .001, Cox and Snell R² = .028, Nagelkerke R² = .063; Model 2, Block 1, Step 1: –2 log likelihood= 289.502, χ²(1) = 9.483, p = .002, Cox and Snell R² = .019, Nagelkerke R² = .042; Model 2, Block 1, Step 2: –2 log likelihood = 285.666), χ²(2) = 13.319, p = .001, Cox and Snell R² = .026, Nagelkerke R²= .058; Model 2, Block 2, Step 1: –2 log likelihood = 278.138), χ²(3) = 20.846, p = .000, Cox and Snell R² = .040, Nagelkerke R² = .090; Model 3, Block 2, Step 1: –2 log likelihood = 276.814), χ²(3) = 22.171, p = .000, Cox and Snell R² = .043, Nagelkerke R² = .096.

Model 3, Block 1 is identical to Model 2, Block 1.

In an effort to control for a number of demographic variables, two additional stepwise logistic regression were performed in which a series of legal (drug offence) and social-demographic measures (age, gender, education, location, and income) were eligible for entry in the first block of predictors. In the first analysis, they were followed by the MRPI-A scale’s total score (Table 4, Model 2). Among the legal and demographic measures, education (OR =.871, 95% CI = [.795, .954]) and marital status (OR = .533, 95% CI = [.284, 1.000]) were accepted into the model in Step 1 and Step 2, respectively. The MRPI Abridged Scale total score was entered in Block 2, indicating a significant contribution to the prediction of recidivism (OR = .373, 95% CI = [.182, .763]) that was independent of education and marital status.

In the final stepwise logistical regression, the five religiosity factors were substituted for the MRPI-A scale total score as potential contributors to the prediction of recidivism in Block 2 (Table 4, Model 3). Only one factor (Religious–Moral Values) was added to the previously included demographic characteristics (education and marital status) and significantly increased the predictive validity of the model (OR = 0.791, 95% CI = [.679, .921]). When comparing Model 3 with Model 1, it is apparent that Religious Practices, in Model 1, was superseded by the demographic measures, education, and marital status, in Model 3, indicating that Religious Practices, but not Religious–Moral Values, was redundant when considering Education and Marital Status. This is not surprising, particularly in light of the fact that education was significantly correlated with Religious Practices (r = .276, p < .01). The same overall pattern of results was obtained when Muslim offenders only were considered.

In sum, religiosity, as measured by the MRPI-A total score (Model 2) or by the Religious–Moral Values factor (Model 3), played a significant role in predicting offender recidivism, independent of various personal and demographic characteristics of the probationers. Findings from the logistic regression also revealed that combining demographic characteristics (education and marital status) with religiosity (BRRI: Abridged total score or the Religious–Moral Values factor score) improved the prediction of probationers’ recidivism (see Table 4, notes), indicating that some particular combination of religiosity factors and demographic background characteristics can better explain probationer recidivism than either domain on its own.

Discussion

The present study investigated the association between religiosity and recidivism, which included both reoffending and violations of probation conditions, in a sample of first-time, adult, nonviolent, community offenders in the Punjab province of Pakistan. The MRPI-A religiosity scale was used to examine the relationship between religiosity and recidivism. This enabled two kinds of investigation, the development and examination of the psychometric characteristics of a revised religiosity measure and an examination of the relationship between religiosity and crime in a highly Muslim culture. The sample was divided into groups defined by religion, gender, marital status, geographical location, and offence type for an examination of various findings by subgroups.

The MRPI-A

Most of the items (11 out of 14) included in the MRPI-A scale were taken either directly or were adapted from the religious personality subscale of the MRPI. Factor analysis revealed that the scale comprised five different, but correlated, religiosity factors. These factors were also examined in relation to recidivism. A psychometric investigation of the MRPI-A scale revealed a moderate degree of internal consistency with this particular sample. The alpha coefficient for the complete scale (.725) met Nunnally’s (1978) acceptability criterion of .70, but this is not considered high. This is not surprising or undesired, given the multidimensionality of religiosity as conceptualized in the original and abridged version of the MRPI. The internal consistency of the five items with factor loadings greater than .40 on Factor 1 (Religious Practices) was actually higher than Cronbach’s (1951) alpha for the complete scale. The remaining four factors did not demonstrate sufficient internal consistency to be considered homogeneous subscales. However, they consisted of only one to three items, and alpha is heavily dependent on the number of items (Schmitt, 1996; Sijtsma, 2007). Their inclusion was also guided by our multidimensional perspective of religiosity, including internal and external religiosity. Further research should seek to examine whether revisions to the items of the MRPI-A scale or the inclusion of more items would generate more internally consistent subscales.

This study supported prior research (e.g., Annalakshmi & Abeer, 2012; Krauss et al., 2007) pertaining to the applicability of the religious personality subscale of the MRPI with followers of different religions and faiths (e.g., Hindus, Muslims, Christians, and Buddhists) in various cultures of Malaysia and India. It also supported the integrity of the MRPI as a well-articulated religiosity measurement model that was based on theoretical propositions and psychometric analysis (Krauss et al., 2006). It accomplished these goals in a couple of ways. First, there were minimal differences between Muslims and the relatively few Christians in the sample in both mean scores and correlations with other measures, thus supporting the contention that it may be applicable to followers of other faiths or at least to Christians. Second, the study offers predictive criterion validity and construct validity to the MRPI-A scale because it correlated (negatively) with subsequent antisocial behavior (criminal recidivism) among Pakistani probationers and because most theoretical and empirical constructions of religiosity link it to the human repertoire of prosocial and antisocial behavior (e.g., Baier & Wright, 2001; Ellison, 1992; Preston, Salomon, & Ritter, 2014; Salas-Wright, Vaughn, & Maynard, 2014).

There was no significant difference in religiosity scores between male and female, urban and rural, married and unmarried groups of adult probationers, nor between Muslim and Christian probationers. However, those convicted of drug-related offences had lower religiosity scores than those convicted of the remaining types of offences combined (i.e., theft, weapon carrying, and miscellaneous offences). This finding corresponds with other previous research (e.g., Baier & Wright, 2001; Benda & Corwyn, 1997; Fernando, Wilson, Staton, & Leukefeld, 2005) that reported differences in religiosity between perpetrators of various types of crime.

This study also contributed to our understanding of the multidimensional nature of religiosity, particularly as it pertains to a highly religious, predominantly Muslim culture. Allowing for correlated factors, 14 items were reduced to five factors in the principal component analysis. As other researchers have found (e.g. Evans et al., 1995), Religious Practice and Religious–Moral Values were the most prominent in terms of number of items (five and three, respectively), variance in the instrument, and correlation with the behavioral outcome, criminal recidivism. They were also highly correlated with each other. A third factor, Importance of Religion (three items), was not correlated with criminal behavior. However, its items generated the highest mean score, indicating that religion was very important to this sample of probationers. The remaining factors may be considered minor, at least with the current sample. The fourth factor, Fundamental Religious Beliefs (two items), was correlated with the first three factors and recidivism, but not incrementally with recidivism. The last factor, Rejection of Nonbeliever (one item), had lower correlations with other factors, was not correlated with recidivism, and appeared to be less important to the current sample as measured by the mean item score. Future research on the MRPI-A might consider eliminating this item.

These findings also speak more generally to religiosity of which there are numerous conceptualizations. One issue in this discussion has focused on whether the construct should be limited to religious practices, and hence be separate from the concept of spirituality, or should encompass religious practices and values. Some researchers have done the former. They reserved the term religiosity for religious practice, such as church attendance, and distinguished it from spirituality, or personal religious beliefs, which is perceived as a separate construct and should be examined separately in relation to prosocial and antisocial behavior (e.g., Good & Willoughby, 2006; Martin, Kirkaldy, & Siefen, 2003; Steinman & Zimmerman, 2004).

In adapting the MRPI-A scale, we elected to do the latter. This is evident in our choice of items and the resulting factor structure of the MRPI-A scale, which is reminiscent of other religiosity measures that include two factors. These factors are often described as “private religiosity” and “public religiosity” (e.g., Kauffman, 1979; Strayhorn, Weidman, & Larson, 1990), a distinction that goes back to ancient Roman times (Dowden, 1992). Other “composite” measures do not make this distinction and may be described as general religiosity scales (e.g., Kerestes, Youniss, & Metz, 2004; King & Furrow, 2004; Regnerus, 2003; Wills, Yaeger, & Sandy, 2003). By conducting an oblique rotation in our principal component analysis of the BRPS: Abridged scale, we allowed for correlated factors. Consequently, the Religious Practice factor was significantly correlated with Religious–Moral Values, both of which were related to recidivism. Use of the Muslim Religiosity-Personality Inventory: Abridged (MRPI-A) also afforded an opportunity to explore theoretical explanations for the relationship between religiosity and crime (see below).

Religiosity and Crime

The correlation and regression analyses demonstrated significant negative associations between overall religiosity and recidivism among adult probationers, indicating that higher religiosity scores were associated with lower rates of reoffending and probation violations among Pakistani probationers. It is noteworthy that this correlation was found among a relatively homogeneous group of low-risk offenders. It is quite likely that the magnitude of the religiosity–crime link would increase among a more diverse sample of offenders in terms of their offence history and severity. Moreover, the relationship was independent of education and marital status. These results were consistent with the broad pattern of the conclusions reported by previous research on a religion–crime relationship across age, gender, and jurisdictions (Baier & Wright, 2001; Benda et al., 2003; Evans et al., 1995; Kerley et al., 2010; Sinha et al., 2007). They also support the general notion that faith/religion may contribute to reductions in criminal activities worldwide among varying economic, social, and cultural settings, including developed and developing countries. Although the relationship between religiosity and recidivism was predictive, one is reminded that it was not demonstrated causally, thus making the previous comment speculative and in need of experimental investigation.

Our research also generated interesting results when we examined the association between religiosity factors and reoffending. Among five various religiosity factors, regression analyses supported only two of them (i.e., Religious Practices and Religious–Moral Values) as they were significantly associated with recidivism among adult probationers (Model 1). Despite a significant relationship between fundamental religious beliefs and recidivism in the sample, as revealed by correlation analysis, the attitudinal dimensions of religiosity, Religious Beliefs (e.g., importance of religion in life, perception of God and belief in the day of judgment, or feeling of shame/repentance), did not produce an independent association with recidivism. This result was contrary to the conclusions reported by other researchers about a positive relationship between conventional beliefs and risk behavior (Sinha et al., 2007; Yeung et al., 2009).

However, these findings tend to support the general notion that the external or behavioral aspects of religion (e.g., involvement in religious activities including church attendance or provision of religious services or participation in religious gatherings) have been consistently related to various criminal outcomes in samples of offender populations (inmates or community offenders) of different ages, gender, and jurisdictional settings (Baier & Wright, 2001; Evans et al., 1995; Johnson et al., 2000; Kerley et al., 2010; Sinha et al., 2007; Tittle & Welch, 1983; Ullrich & Coid, 2011). One plausible explanation of this finding is that major religions/faiths (e.g., Christianity, Islam) command their followers to perform religious activities in a group setting (Clear & Myhre, 1995). This community aspect supports religiosity among followers through socialization, practice, and the reinforcement of religious doctrine, rituals, and values. It is important to note that behavioral dimensions/aspects of religiosity have been frequently examined and validated in preceding research and have been found to have a negative association between religion-related behavior and criminal behavior (Kerley et al., 2010). Based on the correlational findings of the religious beliefs–recidivism relationship in our research, we contend that attitudinal aspects of religion require further investigation in relation to a broad range of criminal actions.

We investigated the explanatory contribution of the five religiosity factors to recidivism in relation to socioeconomic and offence-type variables and found that only education was significantly associated with reoffending along with two religiosity factors (i.e., Religious Practices and Religious–Moral Values). Regression Model 3 demonstrated that Religious Practices did not make a significant contribution to the explanation of reoffending among adult probationers beyond Religious–Moral Values and education. The current analyses suggest that the Religious Practices factor may be related to recidivism because of its significant positive correlation with education (.276), which in turn was related negatively to recidivism (–.131).

It is noteworthy that the variation in recidivism among offenders explained during the regression analyses improved considerably from 6.3% to 9.6% from Model 1 to Model 3, respectively. Moreover, education emerged as a significant factor in Model 2 to understand the risk of reoffending in addition to the two previously mentioned religiosity factors. Therefore, background variables in combination with the religiosity factors may better explain adult probationers’ risk to reoffend. Surprisingly, the background variables including age, monthly income, gender, geographical location, and offence type were not found to be related to reoffending in the regression analysis and in the correlation analyses. Only drug-related offences were shown to have a relationship with recidivism in our sample. These findings were contrary to the conclusions found in previous research (e.g., Evans et al., 1995; Kerley et al., 2010; Sinha et al., 2007) that suggested that age, sex, and income are significant correlates of criminal acts in both youth and adult offender populations.

Theoretical Implications

Despite the fact that religion has been responsible for numerous wars, some estimate 6% to 7% over the last two millennia (Axelrod & Phillips, 2004; Martel, 2012), and continues to spawn violence among religious zealots, its contribution to a healthy disposition and prosocial behavior is well documented. For example, Religious Practices (e.g., church attendance, participation in religious congregation and activity) and Religious–Moral Values are associated with increased levels of hope, well-being, comfort, meaning (Johnson, Larson, & Pitts, 1997; Kerley, Matthews, & Blanchard, 2005) healthy decision making, positive self-esteem (Robinson, Calhoun, & Glaser, 2007), resilience (Annalakshmi & Abeer, 2012), and self-worth (Krause & Ellison, 2007).

The current investigation informs at least two theoretical perspectives about the relationship between religiosity and criminality. First, although a measure of self-control was not included in the current investigation, it is quite possible that Religious Practice (Factor 1) is related to social bond, while Religious–Moral Values (Factor 2) is related to Gottfredson and Hirschi’s (1990) original meaning of self-control. Both factors were correlated with recidivism, and although they were highly correlated with each other, they also contributed independently to the prediction of recidivism (Model 1, Step 2). Moreover, Religious Practice was replaced by education and marital status when demographic variables were introduced into the prediction formula. As social bonds are derived from social connections to individuals and their institutions, it is quite possible that education and marital status constitute sources of social bonds and thus account for the relationship between religious practice and recidivism. Consequently, Hirschi’s more comprehensive conceptualization of self-control appears to offer a more complete account of the relationship between religiosity and crime. Finally, this characterization of religiosity as having two components, both of which are negatively related to criminality, is also reminiscent of an earlier conceptualization of what Allport and Ross (1967) described as internal and external religiosity.

Second, the general personality, cognition, and social learning perspective (e.g., GPCSL) and the personal, interpersonal, and community-reinforcement theory (PIC-R; Bonta & Andrews, 2017) have attached great significance to personal characteristics of individuals such as beliefs, values, attitudes, and practices in efforts to explain their contribution to expressions of criminal acts (Andrews, 1982). Religion and faith comprise one kind of “value system” construct that may influence the cognitive structure and behavior patterns of individuals in a society. In this regard, the findings of the current research support the viewpoints of a prominent theory of the PCC (Bonta & Andrews, 2017) which is based on an integrative approach as it includes GPCSL and PIC-R to explain the variation in criminal behavior. The results of the current research (e.g., the association between religious practices, religious–moral values, and recidivism among adult probationers) are also relevant to the well-established central eight risk/need factors (grounded in GPCSL theory; Andrews et al., 2011).

The prosocial benefits of religiosity, including social networking and the building of social capital, may contribute to crime desistance (see Giordano, Cernkovich, & Rudolph, 2002; Giordano, Longmore, Schroeder, & Seffrin, 2008) via its positive effects on some of the central eight risk/need factors (e.g., adverse circumstances in family life and marital relationships, school and work, leisure, and substance abuse; see Andrews et al., 2004; Bonta & Andrews, 2017). Consequently, many researchers have investigated the role of religion as a protective factor against deviant behaviors, with some having found an empirical relationship between religiosity and crime (Baier & Wright, 2001; Krauss et al., 2012; O’Connor & Perreyclear, 2002; Robinson, 2008; Salas-Wright, Olate, & Vaughn, 2013). Consequently, it is not surprising that religiosity has been nominated for inclusion in offender risk assessments to explain criminal offending beyond other well-documented psychological and social variables (Sinha et al., 2007), even though few theories have explored the complexity of the relationship between religiosity and criminal behavior (Benda, 1995). Although religiosity was correlated with recidivism, it did not contribute incrementally to the prediction of recidivism over and above the LS/CMI (Bhutta & Wormith, 2016). The current investigation hints that this may have occurred because religiosity is manifest in a number of the central eight risk need domains that comprise the LS/CMI, such as education (demonstrated here), peers, and use of leisure time.

Implications for Probation Practice in Pakistan

The findings of the present study on the relationship between overall religiosity, religious factors, and reoffending among adult probationers in the Punjab province are of relevance for criminal justice agencies, in general, and the R&P Department, in particular. POs may use the knowledge about the association of religiosity with recidivism for offenders in two ways.

First, before the release of an offender on probation, a description of an offender’s religiosity in a presentence report (PSR) could enhance the accuracy of the assessment of risk as it would capture important aspects about the offender that could be neglected in traditional risk assessments. It has been previously emphasized that risk assessments focusing exclusively on static risk factors (i.e., overlooking individual beliefs, values, and cultural setting) may increase the probability of a biased prediction of recidivism (e.g., Rogers, 2000). Overprediction of risk for recidivism may lead the court to make biased decisions that are detrimental both to the offender and to society in terms of financial, psychological, and social costs. As suggested by others (e.g., Robinson, 2008; Salas-Wright et al., 2013), the current findings demonstrated that religion is significantly related to reduced reoffending and therefore may be examined as a strength, or protective factor, when considering an offender’s candidacy for community supervision by means of probation or parole, particularly when a structured, empirically validated risk assessment is not available. The findings also support the contention that religiosity influences psychosocial adjustment and individual behavior and should be investigated as part of an offender risk assessment (Robinson et al., 2007; Sinha et al., 2007). Some researchers have suggested focusing on strengths (protective/positive factors) may enhance the efficacy of risk management and intervention programs by facilitating the therapeutic alliance, promoting recovery, serving to motivate clients, and reducing negative cognitions among offenders (e.g., Durnescu, 2012; Ullrich & Coid, 2011).

Second, literature reviews on rehabilitation programs have revealed that religion is not frequently examined as an intervention despite the preponderance of religious programming in corrections and the criminal justice system (Johnson, 2004; O’Connor & Perreyclear, 2002). Clinicians who provide intervention (rehabilitation) programs to offenders should keep in mind the potential for religious affiliation and personal religiosity to contribute to lower rates of reoffending. Although correlational in nature, the fact that drug-related offenders had lower religiosity scores than the remaining probationers may speak to the etiology of drug offending in Pakistan (and elsewhere) and may assist staff in designing and offering treatment for these offenders.

Limitations and Conclusions

The current study has a number of limitations. First, we have examined the association between religiosity and reoffending among adult offenders (above 18 years) from only four districts of the Punjab province of Pakistan. Therefore, the findings of the study may lack generalizability to the population of probationers of Punjab, with its 36 districts, or to the rest of this very diverse country. Moreover, juvenile offenders were excluded and the female sample size was very small. Incarcerated offenders were also excluded from the sample, as it was based on first-time, nonviolent community offenders (probationers) convicted of minor offences (e.g., drug-related offences, theft, and carrying weapons). Second, another limitation of the study was the use of a single measure of recidivism (e.g., termination/cancellation of probation order, which is issued when new criminal charges or violations of the probation order occur), which was the only measure provided to the researchers by the R&P Department. Moreover, the follow-up time for recidivism was quite limited and presumably one of the reasons for the very low recidivism rate. A longer follow-up period, such as a 2-year period, as recommended by Geraghty and Woodhams (2015), would have likely generated a higher recidivism rate and possibly a stronger relationship between religiosity and recidivism.

Third, given the complexity of the religiosity construct, having multiple dimensions related to religion and faith, it is naive to claim that we have precisely assessed the level of religiosity in the current sample. It is no less a challenge than to capture attitudinal, behavioral, and philosophical aspects of a religion from a single religiosity scale (Krauss et al., 2007).

Although the “Religiosity Personality” subscale of the MRPI from which the currently used MRPI-A scale was adapted has established reliability and validity across various cultures (e.g., Hindu, Muslim, Christian, Buddhist) in Malaysia (Krauss et al., 2007) and India (Annalakshmi & Abeer, 2012), the conceptual framework of the religious personality scale (see MRPI; Krauss, Hamzah, Juhari, & Hamid, 2005; Krauss et al., 2005) may have affected the measurement of religiosity differently for Muslim and Christians. Although minimal differences were found between Christian and Muslim probationers on the MRPI-A scale and the scale’s correlation with recidivism, the small sample of Christians prevented a more thorough examination of the scale by type of religion. Moreover, this newly derived version of the MRPI has not been used previously, so caution and further research are suggested. In particular, more prospective research is required to examine the construct of religiosity psychometrically, including the internal consistency of the five subscales, and in relation to criminal behavior in multiple cultures.

Despite these limitations, the present study has contributed to the theoretical and empirical basis of a religion–crime relationship, most of which has been examined among English-speaking offender populations (Ghorbani, Watson, Ghramaleki, Morris, & Hood, 2000). It offers further support to the contention that religiosity is significantly associated with criminal behavior and has done so by examining first-time, non-English-speaking, probationers from a developing Muslim country.

Footnotes

Appendix

Muslim Religiosity-Personality Inventory: Abridged (MRPI-A) and their sources.*

From Allah, Subhanahu Wa Ta’ala (SWT), Most High says in Noble Qur’an, “Ask your Lord for forgiveness and then turn in repentance to Him . . .” (11:3). Also, “O you who believe! Turn (in repentance) to Allah with sincere repentance; Perhaps your Lord will remove your evil from you . . .” (66:8).

*Permission was granted to adapt these items from Steven Kraus, on March 19, 2012. The sources for individual items are not listed in the questionnaire.

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.

Notes

References

Abu Raiya

Pargament

K. I.

Mahoney

Stein

. (2008). A psychological measure of Islamic religiousness: Development and evidence for reliability and validity. The International Journal for the Psychology of Religion, 18, 291-315.

Adamczyk

(2012). Understanding delinquency with friendship group religious context. Social Science Quarterly, 93, 482-505.

Allport

G. W.

Ross

M. J.

(1967). Religious orientation and prejudice. Journal of Personality and Social Psychology, 5, 432-443.

Amer

M. M.

Bagasra

(2013). Psychological research with Muslim Americans in the age of Islamophobia: Trends, challenges, and recommendations. American Psychologist, 68, 134-144.

Andrews

D. A.

(1982). A personal, interpersonal, community-reinforcement perspective on deviant behaviour (PIC-R). Toronto, ON, Canada: Ontario Ministry of Correctional Services. Retrieved from https://www.ncjrs.gov/App/Publications/abstract.aspx?ID=89816

Andrews

D. A.

Bonta

Wormith

J. S.

(2004). The Level of Service/Case Management Inventory: An offender assessment system. User’s Manual. Toronto, ON, Canada: Multi-Health Systems.

Andrews

D. A.

Bonta

Wormith

J. S.

(2006). The recent past and near future of risk and/or need assessment. Crime & Delinquency, 52, 7-27.

Andrews

D. A.

Guzzo

Raynor

Rowe

R. C.

Rettinger

L. J.

Brews

Wormith

J. S.

(2011). Are the major risk/need factors predictive of both female and male reoffending? A test with the eight domains of the Level of Service/Case Management Inventory. International Journal of Offender Therapy and Comparative Criminology, 56, 113-133.

Annalakshmi

Abeer

(2012). Islamic worldview, religious personality and resilience among Muslim adolescent students in India. Europe’s Journal of Psychology, 7, 716-738.

10.

Axelrod

Phillips

(2004). Encyclopedia of wars. New York, NY: Facts on File.

11.

Baier

C. J.

Wright

B. R. E.

(2001). “If you love me, keep my commandments”: A meta-analysis of the effect of religion on crime. Journal of Research in Crime and Delinquency, 38, 3-21.

12.

Bandura

(1977). Social learning theory. Oxford, UK: Prentice-Hall.

13.

Beaver

K. M.

Gibson

C. L.

Jennings

W. G.

Ward

J. T.

(2009). A gene x environment interaction between DRD2 and religiosity in the prediction of adolescent delinquent involvement in a sample of males. Biodemography and Social Biology, 55, 71-81.

14.

Benda

B. B.

(1995). The effect of religion on adolescent delinquency revisited. Journal of Research in Crime and Delinquency, 32, 446-466.

15.

Benda

B. B.

Corwyn

R. F.

(1997). Religion and delinquency: The relationship after considering family and peer influences. Journal for the Scientific Study of Religion, 36, 81-92.

16.

Benda

B. B.

Toombs

N. J.

(2000). Religiosity and violence: Are they related after considering the strongest predictors? Journal of Criminal Justice, 28, 483-496.

17.

Benda

B. B.

Toombs

N. J.

(2002). Two preeminent theoretical models: A proportional hazard rate analysis of recidivism. Journal of Criminal Justice, 30, 217-228.

18.

Benda

B. B.

Toombs

N. J.

Peacock

(2003). Discriminators of types of recidivism among boot camp graduates in a five-year follow-up study. Journal of Criminal Justice, 31, 539-551.

19.

Bhutta

M. H.

(2010). Community based rehabilitation of offenders: An overview of probation and parole system in Pakistan. Pakistan Journal of Criminology, 2, 51-67.

20.

Bhutta

M. H.

(2013). Risk and need assessment of offenders on probation in Lahore (Unpublished doctoral thesis), University of the Punjab, Lahore, Pakistan.

21.

Bhutta

M. H.

Wormith

J. S.

(2016). An examination of a risk/needs assessment instrument and its relation to religiosity and recidivism among probationers in a Muslim culture. Criminal Justice and Behavior, 43, 204-229. doi:10.1177/0093854815604011

22.

Bonta

Andrews

D. A.

(2017). The psychology of criminal conduct (6th ed.). New York, NY: Routledge.

23.

Burkett

S. R.

White

(1974). Hellfire and delinquency: Another look. Journal for the Scientific Study of Religion, 13, 455-462.

24.

Clear

T. R.

Myhre

(1995). A study of religion in prison: The international association of residential and community alternatives. Journal on Community Corrections, 6, 20-25.

25.

Cretacci

M. A.

(2003). Religion and social control: An application of a modified social bond on violence. Criminal Justice Review, 28, 254-277.

26.

Cronbach

L. J.

(1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16, 297-334.

27.

Cullen

(2010). Towards a criminology of religion: Comment on Johnson and Jang. In Rosenfeld

Quinet

Carcia

(Eds.), Contemporary issues in criminological theory and research: The role of social institutions (pp. 151-159). Belmont, CA: Wadsworth.

28.

Dowden

(1992). Religion and the Romans. London, England: Bristol Classical Press.

29.

Durnescu

(2012). What matters most in probation supervision: Staff characteristics, staff skills or programme? Criminology and Criminal Justice, 12, 193-216.

30.

Ellison

C. G.

(1992). Are religious people nice people? Evidence from the National Survey of Black Americans. Social Forces, 71, 411-430.

31.

Evans

T. D.

Cullen

F. T.

Dunaway

R. G.

Burton

V. S.

(1995). Religion and crime re-examined: The impact of religion, secular controls, and social ecology on adult criminality. Criminology, 33, 195-224.

32.

Fernando

Wilson

J. F.

Staton

Leukefeld

(2005). Exploring the type-of-crime hypothesis, religiosity, and spirituality in an adult male prison population. International Journal of Offender Therapy and Comparative Criminology, 49, 682-695.

33.

Geraghty

K. A.

Woodhams

(2015). The predictive validity of risk assessment tools for female offenders: A systematic review. Aggression and Violent Behavior, 21, 25-38.

34.

Ghorbani

Watson

P. J.

Ghramaleki

A. F.

Morris

R. J.

Hood

R. W.

Jr. (2000). Muslim Attitudes Towards Religion Scale: Factors, validity and complexity of relationships with mental health in Iran. Mental Health, Religion & Culture, 3, 125-132.

35.

Giordano

P. C.

Cernkovich

S. A.

Rudolph

J. L.

(2002). Gender, crime, and desistance: Toward a theory of cognitive transformation. American Journal of Sociology, 107, 990-1064.

36.

Giordano

P. C.

Longmore

M. A.

Schroeder

R. D.

Seffrin

P. M.

(2008). A life-course perspective on spirituality and desistance from crime. Criminology, 46, 99-132.

37.

Girard

Wormith

J. S.

(2004). The predictive validity of the Level of Service Inventory–Ontario Revision on general and violent recidivism among various offender groups. Criminal Justice and Behavior, 31, 150-181.

38.

Good

Willoughby

(2006). The role of spirituality versus religiosity in adolescent psychosocial adjustment. Journal of Youth and Adolescence, 35, 41-55.

39.

Gottfredson

M. R.

Hirschi

(1990). A general theory of crime. Stanford, CA: Stanford University Press.

40.

Grasmick

H. G.

Bursik

R. J.

Cochran

J. K.

(1991). Render unto Caesar what is Caesar’s. The Sociological Quarterly, 32, 251-266.

41.

Hamid-uz-Zafar

(1961). Probation in Pakistan; a guide to the Pakistan probation of offenders ordinance, 1960, as implemented in the province of West Pakistan. Lahore, West Pakistan: Government Printing Press.

42.

Hassan

(2007). On being religious: Patterns of religious commitment in Muslim societies. The Muslim World, 97, 437-478.

43.

Heaton

(2006). Does religion really reduce crime? The Journal of Law and Economics, 49, 147-172.

44.

Higgins

P. C.

Albrecht

G. L.

(1977). Hellfire and delinquency revisited. Social Forces, 55, 952-958.

45.

Hirschi

(1969). Causes of delinquency. Berkley: University of California Press.

46.

Hirschi

(2004). Self-control and crime. In Baumeister

R. F.

Vohs

K. D.

(Eds.), Handbook of self-regulation: Research, theory, and applications (pp. 537-552). New York, NY: Guilford Press.

47.

Hirschi

Stark

(1969). Hellfire and delinquency. Social Problems, 17, 202-213.

48.

Jensen

K. D.

Gibbons

S. G.

(2002). Shame and religion as factors in the rehabilitation of serious offenders. Journal of Offender Rehabilitation, 35, 209-224.

49.

Johnson

B. R.

(2004). Religious programs and recidivism among former inmates in prison fellowship programs: A long-term follow-up study. Justice Quarterly, 21, 329-354.

50.

Johnson

B. R.

(2008). A tale of two religious effects: Evidence for the protective and prosocial impact of organic religion. In Kline

K. K.

(Ed.), Authoritative communities: The scientific case for nurturing the whole child (pp. 187-225). New York, NY: Springer.

51.

Johnson

B. R.

Jang

S. J.

(2010). Crime and religion: Assessing the role of the faith factor. In Rosenfeld

Quinet

Carcia

(Eds.), Contemporary issues in criminological theory and research: The role of social institutions (pp. 117-149). Belmont, CA: Wadsworth.

52.

Johnson

B. R.

Jang

S. J.

Larson

D. B.

De Li

. (2001). Does adolescent religious commitment matter? A re-examination of the effects of religiosity on delinquency. Journal of Research in Crime and Delinquency, 38, 22-44.

53.

Johnson

B. R.

Larson

D. B.

De Li

Jang

S. J.

(2000). Escaping from the crime of inner cities: Church attendance and religious salience among disadvantaged youth. Justice Quarterly, 17, 377-391.

54.

Johnson

B. R.

Larson

D. B.

Pitts

T. C.

(1997). Religious programs, institutional adjustment, and recidivism among former inmates in prison fellowship programs. Justice Quarterly, 14, 145-166.

55.

Kauffman

J. H.

(1979). Social correlates of spiritual maturity among North American Mennonites. In Mobe

D.O.

(Ed.), Spiritual well-being: Sociological perspectives (pp. 237-254). Washington, DC: University Press of America.

56.

Kerestes

Youniss

Metz

(2004). Longitudinal patterns of religious perspective and civic integration. Applied Developmental Science, 8, 39-46.

57.

Kerley

K. R.

Copes

Tewksbury

Dabney

D. A.

(2010). Examining the relationship between religiosity and self-control as predictors of prison deviance. International Journal of Offender Therapy and Comparative Criminology, 55, 1251-1271.

58.

Kerley

K. R.

Matthews

T. L.

Blanchard

T. C.

(2005). Religiosity, religious participation, and negative prison behaviors. Journal for the Scientific Study of Religion, 44, 443-457.

59.

King

P. E.

Furrow

J. L.

(2004). Religion as a resource for positive youth development: Religion, social capital, and moral outcomes. Developmental Psychology, 40, 703-713.

60.

Krause

Ellison

C. G.

(2007). Parental religious socialization practices and self-esteem in late life. Review of Religious Research, 49, 109-127.

61.

Krauss

S. E.

Hamzah

A. H.

(2009). The Muslim Religiosity-Personality Inventory (MRPI). Serdang, Malaysia: Universiti Putra Malaysia.

62.

Krauss

S. E.

Hamzah

A. H.

Idris

(2007). Adaptation of a Muslim Religiosity Scale for use with four different faith communities in Malaysia. Review of Religious Research, 49, 147-164.

63.

Krauss

S. E.

Hamzah

A. H.

Ismail

I. A.

Suandi

Hamzah

S. R.

Dahalan

Idris

(2012). Parenting, community, and religious predictors of positive and negative developmental outcomes among Muslim adolescents. Youth & Society, 46, 201-227.

64.

Krauss

S. E.

Hamzah

A. H.

Juhari

Hamid

J. A. B. D

. (2005). The Muslim Religiosity-Personality Inventory (MRPI): Towards understanding differences in the Islamic religiosity among the Malaysian youth. Pertanika Journal of Social Sciences & Humanities, 13, 173-186.

65.

Krauss

S. E.

Hamzah

A. H.

Suandi

Noah

S. M.

Juhari

Manap

J. H.

Mahmood

(2006). Exploring regional differences in religiosity among Muslim youth in Malaysia. Review of Religious Research, 3, 238-252.

66.

Krauss

S. E.

Hamzah

A. H.

Suandi

Noah

S. M.

Mastor

K. A.

Juhari

Manap

(2005). The Muslim Religiosity-Personality Inventory (MRPI)’s religiosity measurement model: Towards filling the gaps in religiosity research on Muslims. Pertanika Journal of Social Sciences & Humanities, 13, 131-145.

67.

Martel

(2012). The encyclopedia of war. Hoboken, NJ: Wiley-Blackwell.

68.

Martin

Kirkaldy

Siefen

(2003). Antecedents of adult well-being: Adolescent religiosity and health. Journal of Managerial Psychology, 18, 453-470.

69.

Maruschak

L. M.

Bonczar

T. P.

(2014). Probation and parole in the United States, 2012 (NCJ 243826). Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics. Retrieved from http://www.bjs.gov/content/pub/pdf/ppus12.pdf

70.

Nunnally

J. C.

(1978). Psychometric theory (2nd ed.). New York, NY: McGraw-Hill.

71.

O’Connor

T. P.

Perreyclear

(2002). Prison religion in action and its influence on offender rehabilitation. Journal of Offender Rehabilitation, 35, 11-33.

72.

Government of Pakistan. (1960). Probation of Offenders Ordinance, 1960. Islamabad, Pakistan: Interior Ministry, Government of Pakistan.

73.

Preston

J. L.

Salomon

Ritter

R. S.

(2014). Religious prosociality: Personal, cognitive, and social factors. In Saroglou

(Ed.), Religion, personality, and social behavior (pp. 149-169). New York, NY: Psychology Press.

74.

Regnerus

M. D.

(2003). Moral communities and adolescent delinquency: Religious contexts and community social context. The Sociological Quarterly, 44, 523-554.

75.

Robinson

(2008). The role of religiosity as a protective factor in the lives of female juvenile offenders (Unpublished doctoral dissertation), University of Georgia, Athens.

76.

Robinson

Calhoun

Glaser

(2007). Exploring religion as a protective for female juvenile offenders. Retrieved from https:///demoiselle2femme.org

77.

Rogers

(2000). The uncritical acceptance of risk assessment in forensic practice. Law and Human Behavior, 24, 595-605.

78.

Salas-Wright

C. P.

Olate

Vaughn

M. G.

(2013). The protective effects of religious coping and spirituality on delinquency results among high-risk and gang-involved Salvadoran youth. Criminal Justice and Behavior, 40, 988-1008.

79.

Salas-Wright

C. P.

Vaughn

M. G.

Hodge

D. R.

Perron

B. E.

(2012). Religiosity profiles of American youth in relation to substance use, violence, and delinquency. Journal of Youth and Adolescence, 41, 1560-1575.

80.

Salas-Wright

C. P.

Vaughn

M. G.

Maynard

B. R.

(2014). Profiles of religiosity and their association with risk behavior among emerging adults in the United States. Emerging Adulthood, 3, 1-18.

81.

Schmitt

(1996). Uses and abuses of coefficient alpha. Psychological Assessment, 8, 350-353.

82.

Sijtsma

(2007). On the use, misuse, and the very limited usefulness of Cronbach’s alpha. Psychometrica, 74, 107-120.

83.

Sinha

J. W.

Cnaan

R. A.

Gelles

R. J.

(2007). Adolescent risk behaviors and religion: Findings from a national study. Journal of Adolescence, 30, 231-249.

84.

Spilka

Hood

R. W.

Hunsberger

Gorsuch

(2003). The psychology of religion: An empirical approach (3rd ed.). London, England: Guilford Press.

85.

Steinman

K. J.

Zimmerman

M. A.

(2004). Religious activity and risk behavior among African American adolescents: Concurrent and developmental effects. American Journal of Community Psychology, 33, 151-161.

86.

Strayhorn

J. M.

Weidman

C. S.

Larson

(1990). A measure of religiousness, and its relation to parent and child mental health variables. Journal of Community Psychology, 18, 34-43.

87.

Tittle

C. R.

Welch

M. R.

(1983). Religiosity and deviance: Toward a contingency theory of constraining effects. Social Forces, 61, 653-682.

88.

Ullrich

Coid

(2011). Protective factors for violence among released prisoners: Effects over time and interactions with static risk. Journal of Consulting and Clinical Psychology, 79, 381-390.

89.

Watts

S. J.

(2018). Examining the relationship between genetics, religiosity, and criminal behavior. Criminal Justice and Behavior, 45, 1192-1212.

90.

Wills

T. A.

Yaeger

A. M.

Sandy

J. M.

(2003). Buffering effect of religiosity for adolescent substance use. Psychology of Addictive Behaviors, 17, 24-31.

91.

Wright

B. R. E.

Caspi

Moffitt

T. E.

Silva

P. A.

(1999). Low self-control, social bonds, and crime: Social causation, social selection, or both? Criminology, 37, 479-514.

92.

Yeung

J. W. K.

Chan

Y.-C.

Lee

B. L. K.

(2009). Youth religiosity and substance use: A meta-analysis from 1995 to 2007. Psychological Reports, 105, 255-266.