Violence and Rehabilitative Process

Journal of Addictions & Offender Counseling • April 2016 • Volume 37 35

© 2016 by the American Counseling Association. All rights reserved.

Received 05/06/15
Revised 07/24/15

Accepted 07/27/15
DOI: 10.1002/jaoc.12014

Self-Efficacy Among Adults in
Substance Abuse Treatment:
The Role of Religious Coping

Amanda L. Giordano, Elizabeth A. Prosek, Sahar Loseu,
Cynthia M. Bevly, Julia Stamman, Citlali E. Molina,
Molly M. Callahan, and Richard-Michael Calzada

In substance abuse treatment, general self-efficacy and religiousness are factors
that may support positive outcomes. The authors surveyed clients receiving
substance abuse treatment (N = 121) and found that religious coping predicted
general self-efficacy scores. Clinical implications are discussed.

Keywords: substance abuse, general self-efficacy, religious coping

Bandura (1977) defined self-efficacy as the belief that “one can successfully
execute the behavior required to produce the outcomes” (p. 193). According
to Bandura (1999), one’s belief that he or she can successfully reach a desired
outcome is necessary to initiate action and to persist in light of adversity or
challenges. Researchers have found that the construct of self-efficacy relates
to many behaviors such as exercise (Luszczynska, Gibbons, Piko, & Tekozel,
2004), oral self-care (Schwarzer, Antoniuk, & Gholami, 2015), and persistence
in college (Liao, Edlin, & Ferdenzi, 2014). With particular relevance to the
current study, previous researchers have also found associations between self-
efficacy and substance abuse (Connors, Tonigan, & Miller, 2001; Coon, Pena,
& Illich, 1998; Kuusisto, Knuuttila, & Saarnio, 2011); generally, higher levels
of self-efficacy in specific domains, such as alcohol abstinence, corresponded
to lower levels of substance abuse and better substance abuse treatment out-
comes. In addition to empirical research, previous authors have supported
the relationship between self-efficacy and substance abuse with theoretical
propositions: Bandura (1997) noted that an increase in self-efficacy will likely
lead to sobriety among problem drinkers. With regard to relapse, Bandura
(1999) purported that those with high self-efficacy may conceptualize a brief
return to substance use as a slip, rather than failure, and recommit to sobriety.
Therefore, self-efficacy is an important construct in the exploration of substance
abuse treatment and sustained abstinence from drugs and alcohol.

Amanda L. Giordano, Elizabeth A. Prosek, Sahar Loseu, Cynthia M. Bevly, Julia Stamman, Citlali E.
Molina, Molly M. Callahan, and Richard-Michael Calzada, Department of Counseling and Higher
Education, University of North Texas. Correspondence concerning this article should be addressed
to Amanda L. Giordano, Counseling Program, Department of Counseling and Higher Education,
University of North Texas, College of Education, 1155 Union Circle #310829, Denton, TX 76203
(e-mail: [email protected]).

36 Journal of Addictions & Offender Counseling • April 2016 • Volume 37

Currently, there are two distinct ways to measure self-efficacy: behavior-
specific or general. With regard to substance abuse, numerous researchers
have studied behavior-specific self-efficacy, particularly related to absti-
nence and refusal behaviors, and have demonstrated a strong relationship
between self-efficacy and treatment outcomes (Greenfield, Venner, Kelly,
Slaymaker, & Bryan, 2012; Kadden & Litt, 2011). For example, in a study of
clients in residential treatment for substance use disorders, Ilgen, McKellar,
and Tiet (2005) found abstinence self-efficacy to be the strongest predictor
of sustaining abstinence for 1 year posttreatment. Abstinence self-efficacy
referred to participants’ belief that they will remain completely abstinent
1 year after discharge from the program. In addition, findings from Project
MATCH, the multisite, national research project, revealed abstinence self-
efficacy, as measured by the Alcohol Abstinence Self-Efficacy scale (AASE;
DiClemente, Carbonari, Montgomery, & Hughes, 1994), as a predictor of
days abstinent among adults in outpatient and aftercare treatment for
substance abuse (Connors et al., 2001).

Although behavior-specific self-efficacy is recommended by Bandura
(1997), several researchers have investigated general self-efficacy, encom-
passing an individual’s global sense of competence to address stressful or
adverse situations (Luszczynska, Scholz, & Schwarzer, 2005; Schwarzer &
Jerusalem, 1995). Researchers posited that self-efficacy is a homogeneous
construct (Schwarzer, Mueller, & Greenglass, 1999). Supporters of global
self-efficacy have recommended that researchers consider studying general
self-efficacy when interested in multiple obstacles or stressors in various
domains (Luszczynska et al., 2005). Given the multitude of stressors and
barriers faced by those in recovery from substance abuse, such as damaged
relationships, financial stress or strain, health issues, and guilt or shame
stemming from past experiences, general self-efficacy seems appropriate. A
sense of efficacy regarding the ability to refuse or abstain from substances
may be too narrow and, in fact, may not be the precipitator of relapse.
Indeed, Bandura (1997) suggested that “problem drinkers are driven to
heavy drinking more by a low sense of efficacy to manage aversive emo-
tional states” (p. 359). Therefore, higher levels of general self-efficacy may
allow a recovering individual to address a variety of stressors and therefore
reduce the need to cope with substances.

Previous researchers have investigated the relationship between general
self-efficacy and substance abuse, finding support for an inverse relationship.
Specifically, among adult men, Mittag and Schwarzer (1993) determined that
those with low general self-efficacy had higher alcohol consumption than
those with high general self-efficacy. Among individuals with HIV, Corless
et al. (2012) found those with low general self-efficacy engaged in higher
levels of substance abuse. In addition, Hassel, Nordfjærn, and Hagen (2013)
described a significant, inverse relationship between general self-efficacy
and illicit drug use as well as alcohol use. Moreover, Nordfjærn, Hole, and
Rundmo (2010) reported that psychological distress, the strongest predictor

Journal of Addictions & Offender Counseling • April 2016 • Volume 37 37

of substance abuse, was significantly and inversely related to general self-
efficacy. Therefore, given the established relationship between general
self-efficacy and substance use, it seems imperative to better understand
predictors of general self-efficacy among substance abusing clients. One
potential predictor is religiousness and individual variables pertaining to
religious identity.

Religiousness, Spirituality, and Self-Efficacy

Over the past few decades, the construct of spirituality has emerged as
distinct from religion. Rather than polar opposites, however, religion
and spirituality frequently are integrated constructs (Pargament, 1997).
Spirituality, which often is conceptualized as one’s personal experience
of beliefs, values, and transcendence, may exist with or without religion,
which often refers to institutionalized practices and communal traditions
(Pargament, 1997; Zinnbauer et al., 1997). Although differences exist be-
tween the constructs of spirituality and religion, significant overlap and
integration also exist; therefore, researchers on these topics often explore
elements of both phenomena.

Previous researchers have provided support for the associations between
religious/spiritual elements and various types of self-efficacy. For example,
in a sample of 144 undergraduate students, Duffy and Blustein (2005)
determined that intrinsic religiousness, defined by internally motivated
reasons for one’s religiousness, and spiritual awareness, defined by one’s
awareness of God and the nature of one’s connection to God, significantly
predicted career decision self-efficacy. In a study of African American
youth and young adults, researchers found that substance refusal efficacy
positively correlated with public and private religiosity (Nasim, Utsey,
Corona, & Belgrave, 2006). In addition, refusal efficacy, or one’s confidence
in the ability to refrain from using substances in stressful situations, medi-
ated the relationship between private religiosity (i.e., personal beliefs and
an individual’s values pertaining to a Higher Power) and several forms
of substance use (Nasim et al., 2006). Researchers have also determined
that religiousness is related to general self-efficacy. Holt, Roth, Clark, and
Debnam (2014) found that among a sample of African American adults,
those with higher levels of religious beliefs (i.e., private participation in
religious behaviors and personal religious beliefs) had higher levels of
general self-efficacy. Therefore, previous researchers have suggested that
religiousness correlates with self-efficacy, as measured in various ways.

One aspect of religiousness warranting further examination with regard to
self-efficacy is religious coping. Pargament, Smith, Koenig, and Perez (1998)
suggested that when faced with negative life circumstances, religion can
be used as a positive coping strategy (perceiving God as benevolent and a
source of support) or a negative coping strategy (perceiving God as puni-
tive and punishing). Although related, positive religious coping (PRC) and

38 Journal of Addictions & Offender Counseling • April 2016 • Volume 37

negative religious coping (NRC) are distinct dimensions of the multifaceted
construct of religious coping (Pargament et al., 1998). Previous researchers
have found small, positive correlations between PRC and NRC; however,
the low correlations (i.e., r = .17 and r = .18) and discriminant validity
confirmed that they are unique constructs, rather than opposing ends of a
continuum, which individuals may use simultaneously (Pargament et al.,
1998). Indeed, Pargament et al. (1998) stated that rather than using only
one religious coping method at a time, individuals use religious coping
“in combination with each other” (p. 720) when faced with negative life
events. Both PRC and NRC may be associated with the construct of self-
efficacy. For example, higher levels of PRC were found to associate with
higher levels of psychological growth resulting from a stressor (Pargament
et al., 1998). In addition, in a qualitative study of female intimate partner
violence survivors, researchers identified a theme of self-efficacy emerging
from the use of spiritual coping mechanisms including “experiencing God
as a lifeline for survival” (Drumm et al., 2014, p. 389). This finding supports
the notion that PRC may predict self-efficacy. With regard to NRC, Smither
and Walker (2015) found that core self-evaluation (a construct including
generalized self-efficacy among other factors) stood in a significant negative
association with the perception that God is punitive. Therefore, both PRC
and NRC may be significant predictors of general self-efficacy.

Purpose of the Current Study

Given the extant research findings on general self-efficacy as a positive fac-
tor in substance abuse treatment and recovery, as well as the relationship
between self-efficacy and religious elements, we developed the present
study to determine the predictive validity of religious coping on general
self-efficacy. Specifically, we sought to address the following research
question: Do levels of PRC and NRC significantly explain the variance
of general self-efficacy among a sample of adults in outpatient substance
abuse treatment? In view of previous findings, we hypothesized that both
PRC and NRC would significantly and uniquely explain the variance of
general self-efficacy, with PRC predicting more general self-efficacy and
NRC predicting less.

Method

Procedure

After receiving institutional review board approval, we identified potential
substance abuse treatment centers to serve as the sampling population for
the study. We used convenience sampling methods and selected substance
abuse counseling groups based on size, level of treatment, and time of day
to provide a greater chance for adequate sample distribution. Inclusion cri-
teria for selection included that treatment centers (a) were nonfaith based,

Journal of Addictions & Offender Counseling • April 2016 • Volume 37 39

(b) used intensive outpatient programs, (c) and provided group counseling
for adult clients. We administered 127 paper surveys to group members in
13 separate outpatient groups within three treatment centers. Participants
who completed the survey received a $5 gift card to a local restaurant. We
collected 125 completed survey packets, with a response rate of 98.4%. We
removed four surveys because of missing data, particularly in regard to
age, because we could not verify the participants met the inclusion criteria
of age 18 years. Therefore, 121 participants formed the final sample.

Participants

Participants included adults in outpatient substance abuse treatment centers
in a southwestern region of the United States. Two inclusion criteria for
the study included a minimum age of 18 years and current involvement
with outpatient substance abuse treatment. Participant groups were equal
among those mandated to treatment (47.9%, n = 58) versus those volun-
tarily receiving treatment (47.9%, n = 58), with five individuals choosing
not to provide this information. The mean age of the participants was
35.04 years old (SD = 11.60). With regard to gender, 45.5% (n = 55) of par-
ticipants identified as women, 53.7% (n = 65) identified as men, and one
participant (0.8%) identified as transgender. Racial/ethnic identification
was as follows: 1.7% (n = 2) Asian/Pacific Islander, 29.8% (n = 36) Black/
African American, 20.7% (n = 25) Latino/Hispanic, 1.7% (n = 2) multiracial,
2.5% (n = 3) Native American, 43.0% (n = 52) White, and 0.8% (n = 1) other.
Participants reported sexual orientation: 7.4% (n = 9) of participants identi-
fied as bisexual, 6.6% (n = 8) identified as gay, 85.1% (n = 103) identified as
heterosexual, and one participant (0.8%) did not report sexual orientation.
With regard to religious affiliations, participants selected from a list of 10
potential options in which 0.8% (n = 1) identified as Buddhist, 9.9% (n =
12) identified as Christian Catholic, 57.0% (n = 69) identified as Christian
Protestant, 2.5% (n = 3) identified as practicing New Age spirituality, 6.6%
(n = 8) identified as no religious affiliation, 19.8% (n = 24) identified as
spiritual but not religious, and 3.3% (n = 4) identified as other. Some per-
centages may not total 100 because of rounding.

Measures

Brief Religious Coping Scale. The Brief Religious Coping Scale (RCOPE;
Pargament, Feuille, & Burdzy, 2011; Pargament et al., 1998) is composed
of 14 items assessing religious coping. More specifically, the Brief RCOPE
contains two subscales, one that measures PRC and one that measures
NRC in response to life stressors. The subscales are measured on a 4-point
Likert-type scale, ranging from 1 (not at all) to 4 (a great deal). An example
of a PRC scale item is, “Sought God’s love and care,” and an example of
an NRC scale item is, “Questioned God’s love for me” (Pargament et al.,
1998). Because the scale specifically refers to a Deity, it is conceptualized
as a religious coping scale; however, because many individuals express

40 Journal of Addictions & Offender Counseling • April 2016 • Volume 37

their spirituality through religion (Pargament, 1997), it may also represent
spiritual coping for some participants. In previous research, Giordano et
al. (2015) used the Brief RCOPE and attained a Cronbach’s alpha level of
.92 for the PRC scale and .81 for the NRC scale, providing evidence of
strong internal consistency. With regard to validity, Pargament et al. (2011)
reviewed current literature and determined that PRC scores related to
spiritual health and psychological well-being whereas NRC scores related
to poor psychological functioning. In the current study, scores on the PRC
subscale had a Cronbach’s alpha level of .93; scores on the NRC subscale
had a Cronbach’s alpha level of .87. These subscale scores demonstrated
strong reliability evidence above the minimum standard level of .70 for
social science research (Heppner & Heppner, 2004).

General Self-Efficacy Scale. We used the General Self-Efficacy Scale (GSE;
Schwarzer & Jerusalem, 1995) in this study to measure self-efficacy. The
original GSE scale was constructed in German and has since been trans-
lated to over 28 languages (Luszczynska et al., 2005). Respondents rated
the 10 items on a 4-point Likert-type scale ranging from 1 (not at all true)
to 4 (exactly true). Example items include “I can always manage to solve
difficult problems if I try hard enough” and “I can remain calm when fac-
ing difficulties because I can rely on my coping abilities.” The GSE score
is obtained by calculating the sum of all item scores, with 10 representing
the lowest score and 40 representing the highest. Scores on the GSE scale
demonstrated strong evidence for internal consistency (Cronbach’s α = .86)
in a meta-analysis (Scholz, Doña, Sud, & Schwarzer, 2002). The test–retest
reliability based on a 2-year interval ranged from .47 for men to .63 for
women (Schwarzer & Jerusalem, 1995). In addition, researchers supported
the scale’s validity with positive correlations between GSE and self-esteem,
optimism, and internal control (Schwarzer & Jerusalem, 1995). Scores on the
GSE scale indicated strong reliability evidence (α = .81) in the current study.

Data Analysis

Before data analysis, we used the G*Power software program to calculate
the necessary sample size (Faul, Erdfelder, Buchner, & Lang, 2009). We
chose to specify a medium effect size for a priori analyses on account of
previous findings on related constructs (Hassel et al., 2013). Our study
required 68 participants with a moderate effect size (f 2 = .15), a power of
.80, an alpha level of .05, and two tested predictors. Because this study was
part of a larger research project, we sought to collect data from 125 adults
in substance abuse treatment to perform all necessary analyses and account
for incomplete surveys. Upon initial examination of the data, we assessed
whether or not the data met the assumptions of multivariate analyses. A
correlation matrix revealed modest correlations between study variables.
Thus, multicollinearity was not a concern. The data met assumptions of
independence and linearity. In addition, all skewness and kurtosis coef-
ficients were within the range of plus or minus one, with the exception

Journal of Addictions & Offender Counseling • April 2016 • Volume 37 41

of NRC, which was slightly skewed (1.10). Upon visual inspection of the
scatterplot of standardized predicted and residual values, we determined
that the data met the assumption of homoscedasticity and, therefore, we
proceeded with the multiple regression analysis. We used a multiple re-
gression analysis to assess the unique contribution of each predictor (PRC
and NRC) on GSE. Because PRC and NRC are distinct yet connected (Par-
gament et al., 1998), a multiple regression allowed for the examination of
the predicted change in GSE resulting from each form of religious coping
while holding the other constant.

Results

The means and standard deviations of study variables were as follows:
GSE (M = 30.31, SD = 4.57), PRC (M = 21.45, SD = 5.96), and NRC (M =
12.06, SD = 5.31). To address the research question examining the amount
of variance in general self-efficacy scores accounted for by PRC and NRC,
we used a simultaneous multiple regression. PRC and NRC scores served
as the independent variables, and general self-efficacy scores served as the
dependent variable. The results of the regression analysis were significant,
R2 = .07, adjusted R2 = .06, F(2, 118) = 4.61, p = .01, accounting for 7.2% of the
variance. According to Cohen (1992), this effect size falls between small (.02)
and medium (.15). Examination of standardized beta coefficients revealed
both PRC, B = .18, β = .24, t = 2.61, p = .01, 95% CI [.04, .32], and NRC, B
= –.16, β = –.18, t = –2.02, p < .05, 95% CI [–.31, –.00], uniquely accounted for variance explained. The direction of the beta values, which were con- firmed by the correlation coefficients, indicated that higher levels of PRC and lower levels of NRC significantly predicted general self-efficacy scores among adults in outpatient substance abuse treatment, thereby supporting our hypothesis. Discussion Beyond the established associations between GSE and lowered substance use (Corless et al., 2012; Hassel et al., 2013; Mittag & Schwarzer, 1993), GSE relates to other desirable outcomes such as quality of life (Ponizovsky et al., 2010), active coping and positive reframing (Luszczynska et al., 2005), and proactive attitudes (Albion, Fernie, & Burton, 2005). In light of the potential benefits associated with GSE, it is important to identify predictors of GSE. Such findings might inform substance abuse treatment practices. In the current study, we sought to determine the amount of variance of GSE explained by PRC and NRC among adults in outpatient substance abuse treatment. Religious elements have long been recognized as protective fac- tors against substance abuse (Desmond, Ulmer, & Bader, 2013; Fletcher & Kumar, 2014; Mason, Schmidt, & Mennis, 2012; Stoltzfus & Farkas, 2012), yet they also may relate to self-efficacy. Indeed, our findings indicated that, 42 Journal of Addictions & Offender Counseling • April 2016 • Volume 37 after accounting for interrelationships among the predictor variables, the regression model (containing both PRC and NRC) significantly predicted GSE. Specifically, PRC uniquely predicted higher levels of GSE whereas NRC uniquely predicted lower levels, which supported our hypothesis. Religious coping explained slightly more than 7% of the variance in GSE, a small to moderate effect size. These findings were similar to those attained by previous researchers who found associations between aspects of religiousness and self-efficacy (Duffy & Blustein, 2005; Holt et al., 2014). Our study served to expand these findings by supporting the relationship between religious elements and self-efficacy among clients with substance use disorders. In addition, our results confirmed the distinct effects of PRC and NRC on self-efficacy. Previous researchers supported the notion that spiritual coping responses and PRC are linked to self-efficacy and psychological growth (Drumm et al., 2014; Pargament et al., 1998). In the same way, PRC predicted general self-efficacy in our regression model. One possible explanation for this find- ing is that those with a secure connection to the Divine feel more capable to execute tasks successfully with the support of a Higher Power. Furthermore, previous researchers have linked punitive perceptions of God (related to NRC) to lower self-evaluation, including generalized self-efficacy (Smither & Walker, 2015). Similarly, the results of our regression model supported the negative association between NRC and GSE. The utilization of NRC may decrease self-efficacy by causing individuals to question their ability to be successful if the Divine is “punishing” or working against them. Thus, an individual’s perception of God and function of religiousness during nega- tive life circumstances may serve to empower or paralyze the individual in his or her quest to be successful in a variety of tasks. Implications for Counselors These findings have several implications for clinicians working in substance abuse treatment settings, such as emphasizing the importance of assessment for religious/spiritual elements of clients’ identities. Rather than relying on demographic items on an intake form, more thorough assessment is needed to understand how clients use religiousness, particularly when faced with challenging circumstances. Pargament (1997) defined religion as “a search for significance in ways related to the sacred” (p. 32) and em- phasized that clients can use religion as a way to positively or negatively cope with challenges. For clients who identify interest or experience in such matters, an exploration of their religious belief system—or absence of religious belief system—may be useful at both intake and throughout the counseling process. This focus on assessment is supported by the spiritual competencies crafted by the Association for Spiritual, Ethical, and Religious Values in Counseling (ASERVIC; Cashwell & Watts, 2010). Competency 10 states that counselors Journal of Addictions & Offender Counseling • April 2016 • Volume 37 43 work to “understand a client’s spiritual and/or religious perspective by gathering information from the client and/or other sources” (Cashwell & Watts, 2010, p. 5). A comprehensive understanding of the client’s religious perspective, when applicable, can assist in building rapport, constructing accurate client conceptualizations, determining interventions, and forming appropriate treatment goals. Therefore, it may behoove counselors work- ing with an addictions population to examine both the nature of a client’s religiousness and whether clients use religious elements in response to difficulties. Counselors may use formal measures such as the Brief RCOPE (Pargament et al., 1998), Quest Scale (Batson & Schoenrade, 1991), or Daily Spiritual Experience Scale (Underwood & Teresi, 2002) to assess religious/ spiritual elements. In addition, counselors can invite clients to construct spiritual life maps (Hodge, 2005), in which clients depict significant events on their spiritual journeys. Finally, counselors can use informal assessment strategies throughout the counseling process by reflecting, exploring, and assisting clients in their understanding of their religious/spiritual values and identities. Another implication is the importance of supporting clients’ use of PRC, given our findings that PRC was related to higher levels of GSE, whereas NRC was related to lower levels of GSE. Pargament et al. (1998) stated that PRC results from a secure connection with God and consists of coping strategies, such as seeking spiritual support, collaboration, and benevolent reappraisals of God. NRC, conversely, results from a tenuous connection with God, spiritual struggle, and punishing reappraisals of God (Pargament et al., 1998). One way to support clients’ use of PRC may be through an attachment theory framework. Researchers have linked God images and God concepts to childhood and adult attachment styles (Dickie et al., 1997; Noffke & Hall, 2007). Moriarty, Hoffman, and Grimes (2006) posited that counselors can assist in changing attachment styles and related God images by (a) creating a secure bond between counselor and client; (b) responding in new, healthy ways to client transference; and (c) identifying maladaptive attachment styles to increase client awareness of relational patterns. In this way, clients can model attachment styles after the therapeutic alliance and extend these patterns to other relationships (Moriarty et al., 2006). By aiding in the formation of secure attachment styles, counselors may help clients assess and reconstruct their religious attachment to foster PRC. Another means to support clients’ use of PRC may involve referrals to 12-step support groups, such as Alcoholics Anonymous (AA). The AA fel- lowship does not affiliate with any religious denomination but does identify as a spiritual program. For clients who are open to the spiritual nature of 12-step programs, AA may offer a helpful, adjunct resource for their recovery process. Through the steps and traditions of AA, members come to rely on a Higher Power and find support through a secure connection with the Divine. The third step of the 12 steps of AA describes surrendering to “the care of God as we understood Him” (Alcoholics Anonymous World 44 Journal of Addictions & Offender Counseling • April 2016 • Volume 37 Services Inc., 2001, p. 59), emphasizing that the God of their understanding is a supportive, caring resource, rather than a punitive, condemning judge, thus fostering PRC rather than NRC. In addition, the Big Book of AA states, The central fact of our lives today is the absolute certainty that our Creator has entered into our hearts and lives in a way which is indeed miraculous. He has commenced to accomplish those things for which we could never do by ourselves. (p. 25) This description of the nature of the Divine may augment PRC levels by creating secure, rather than tenuous, connections with a Higher Power. In- deed, many members of AA have described experiencing a gradual spiritual awakening, after …