Self-Efficacy

Abstract

Self-efficacy in academic settings is an established correlate of educational accomplishments with relevance beyond the classroom. It is a socially created propensity to view oneself as capable of responding to a range of life contingencies. We measure shifts in self-efficacy within prison-based courses that are modeled after the Inside–Out Prison Exchange Program. Courses include college students (outside) and people who are incarcerated (inside) learning together in a prison classroom. Inside students report lower levels of self-efficacy at Time 1 and an increase in self-efficacy by Time 2. Outside student levels of self-efficacy remain the same across time.

Keywords

self-efficacy Inside–Out Prison Exchange Program service learning

Self-efficacy is as a socially constructed individual attribute (Pajares & Miller, 1995), described as a complex and malleable construct (Gist & Mitchell, 1992) rather than a personality trait (Pajares & Miller, 1995). It is associated with social cognitive theory, which postulates a triadic, reciprocal, causal relationship among individuals’ social environments, behaviors, and cognitions (Bandura, 1986). In addition, self-efficacy has important implications for behaviors in a variety of life domains (Grabowski, Call, & Mortimer, 2001, p. 164) throughout the life course. Research indicates that sources of self-efficacy emerge from several important social contexts (e.g., family, recreation, faith traditions, and beliefs; Staples, Schwalbe, & Gecas, 1984), with most focusing on processes that unfold in the workplace or educational settings.

In general, self-efficacy involves “beliefs in one’s capabilities to mobilize the motivation, cognitive resources, and courses of action needed to meet given situational demands” (Gist & Mitchell, 1992, p. 184), and is independent from other elements of the self-concept such as self-esteem or a sense of self-worth and value (Staples et al., 1984). In academic contexts, self-efficacy “refers to subjective convictions that one can successfully carry out given academic tasks at designated levels” (Bong, 2004, p. 288).

As a psychological orientation, self-efficacy is a potent contributor to educational attainment, occupations pursuits, and career outcomes (Staples et al., 1984; Zajacova, Lynch, & Espenshade, 2005). Across these important, interconnected areas, “people initiate and persist in activities that they feel capable of conducting successfully; they tend to avoid those which they feel unable to carry out” (Grabowski et al., 2001, p. 164). Among adult learners, academic self-efficacy has a strong, positive influence on college credits and grade point average (GPA) in the first year of college, net of other relevant variables (e.g., stress levels, high school grades, etc.; Zajacova et al., 2005). In the workplace, the experience of self-efficacy is systematically shaped by occupational positions that differ in terms of class and status-related variables (e.g., power, authority, and freedom to exercise judgments and creativity). Such structural features of the workplace thus impact overall self-esteem through the intervening factor of self-efficacy (Staples et al., 1984).

The descriptive study presented here highlights self-efficacy and whether it is influenced during the time frame of a college course held within a prison. The courses in this study are semester-long, total engagement immersion experiences that take place in prisons. Per course, about half of the students are people who are incarcerated (inside) and the other half college students (outside). Course pedagogy is derived from the Inside–Out Prison Exchange Program (Pompa, 2002) that involves weekly templates of engagement that frame the student learning experience (Allred, 2009). Weekly classes are structured to enhance dialogic, collaborative, critical conversations, and include written reflections about crime, justice, and other relevant social issues (Pompa & Crabbe, 2004).

These and other inside–out courses have been described as “transformative learning experiences” that “invite individuals to take leadership in addressing” the issues, policies, and topics studied (Pompa & Crabbe, 2004). As a form of human agency, evidence that self-efficacy emerges from or is sustained within this learning context may highlight one way transformation manifests within individual students. In addition, social cognitive theory suggests (Bandura, 2000) that contexts that give rise to self-efficacy may also contribute to the emergence of another form of human agency, namely collective agency or efficacy (Sampson, Raudenbush, & Earls, 1997; that is, class members shared beliefs in their capacity to be change agents). To be certain, a group (or class) characteristic of this type is independent of members’ individual notions of personal agency (Bandura, 2000), but collective agency is an unlikely outcome without some threshold of self-efficacy within members. Indeed, “a collective system with members plagued by self-doubts about their capabilities to perform their roles will achieve little” (Bandura, 2000, p. 77).

We measure self-efficacy at two time periods, and evaluate inside and outside student data combined and separately. We pursue this analytic strategy given the different daily living contexts that characterize the lives of our student groups. The inside students live either in a prison or a reentry program where they encounter more constraining environmental influences compared with the local context of their outside counterparts. Because self-efficacy emerges from the features that distinguish these diverse environments, we expected differences in self-efficacy levels initially and across time.

Literature Review

Self-Efficacy as a Social Construction

Over the life course, self-efficacy is a highly influential component of the self-concept that is capable of impacting life paths (Bandura, 1990). Our levels of self-efficacy depend less on our actual experiences with success and failure in specific life domains, although these are important. Rather, this self-referent process is influenced more so by our interpretations of our prior accomplishments, verbal feedback provided by significant others, and vicarious learning (Stevens et al., 2004).

Further, people of different age groups assign different psychological meanings to social experiences, life events, and personal attributes (Rosenberg & Pearlin, 1978). For example, among adults, social class constitutes an achieved status, but for children, social class represents an ascribed status. For this reason, in part, we tend to observe a strong relationship between social class and self-esteem (and other self-concept constructs) with increasing age (Gecas & Seff, 1989).

Within groups of relatively similar ages, social positioning (e.g., social class position) impacts self-evaluations in general and in task-specific domains (Gecas & Seff, 1989; Mortimer & Lorence, 1979; Staples et al., 1984). That is, social structural location plays a pivotal role in the processes of self-evaluation and interpersonal comparisons. While there are myriad pathways though which the positioning may affect the values and psychosocial characteristics of individuals (Staples et al., 1984), a key pattern has emerged. Where significant positions in daily living environments afford opportunities for autonomous action (e.g., self-direction, creativity), individuals are more likely to make self-attributions of competence because the structural requirements of the context make effectual action possible.

Understood in this manner, self-efficacy is a function, in part, of the relatively immediate, situational context (Bong, 2004) and is best regarded as task specific rather than a global personality characteristic (Jackson, 2002). Relevant for the focus of this study, the pliability of this self-evaluation has been observed in educational contexts where instructor interventions were evaluated for their personal efficacy-enhancing impacts (Jackson, 2002). And while such empirical results indicate the malleability of this component of the self-concept, transitions that are induced by workplace or pedagogical adjustments are likely to be short-lived where there remains a gap between individual’s objective prospective or current capability and this evaluative component of the self. For example, Gist & Mitchell (1992) examine worker sales activity in conjunction with self-efficacy levels. Intentional efforts to minimize debilitating beliefs that may impact worker performance culminated in increased performance. Yet, there were limits to adjusted conditions effects on self-efficacy levels. They found a human capital linkage with self-efficacy levels: when baseline levels of self-efficacy were relatively low and an inaccurate reflection of performance capabilities, they observed increases in self-efficacy that were stronger and of longer duration. When baseline levels were relatively low and corresponded to an accurate reflection of performance capabilities, changes in self-efficacy were more modest and transient.

In sum, social cognitive theorists (e.g., Bandura, 1977) indicate that knowledge and skill sets are relevant for whether and how we respond to daily challenges, adversities, failures, or other situations that may expose our relative inadequacies. In addition, self-efficacy beliefs are an essential influence in this triadic relationship of behavior, cognitions, and social situations (Bandura, 1977). Such beliefs influence people’s endeavors and successes because they affect environments and situations that people choose. All other things being equal, people prefer environments in which they feel competent and avoid those in which they feel inadequate. When facing normal life challenges, people with high self-efficacy beliefs are more likely to expend greater effort, remain more problem focused, and persist for longer periods of time than those with low self-efficacy (Jackson, 2002). Further, social cognitive theory describes a dynamic between individuals and the groups they populate. With regard to self-efficacy, Bandura describes its presence among members as a necessary, but not sufficient precursor to the group property of collective agency or efficacy (Bandura, 2000).

In the context of a prison culture, normal life challenges that may thwart self-efficacy among incarcerated individuals are of a different form and may assume a different meaning compared with those encountered by the civilian general population or traditional college students. Most concur, for example, that prisons are not healthy places for several reasons (World Health Organization [WHO], 2007). Often prisons and “jails are noisy, chaotic, and violent places” (Kuhlman & Ruddell, 2005), and the prevailing conditions of stress and threats to personal wellbeing are unmatched in situations outside the prison walls (Vega & Silverman, 1988). In addition, prisons tend to be managed and structured for the purposes of limiting the personal comforts and liberties of those within them (Banks, 2003). Residents of prisons are among the most stigmatized, othered groups in our society (Kaposy & Bandrauk, 2012; Schnittker & John, 2007). Last but not least, their occupants tend to have lower educational backgrounds compared with the general population (Bureau of Justice Statistics [BJS], 2003). Within this context, it is not a stretch to question the chances for creating an enclave to enhance a sustainable sense of self-efficacy, when the total institution presents as an inherent conflict with the requisites of academic culture (Banks, 2003).

Service Learning and Student Outcomes in Correctional and Noncorrectional Settings

The National Service-Learning Clearinghouse (2007) defines service learning as “a form of experiential education where learning occurs through a cycle of action and reflection as students work with others through a process of applying what they are learning to community problems and, at the same time, reflecting upon their experience as they seek to achieve real objectives for the community and deeper understanding and skills for themselves” (p. 2). Service learning is designed to promote teamwork while offering opportunities to create strategies that address complex problems in complex situations. Ideally, it enhances critical thinking, promotes learning through active participation, and fosters a sense of caring for others (Starks, Harrison, & Denhardt, 2011). It also involves relationships. . .” A relationship that is based on equality and collaboration. From such a perspective, service is seen more as an act of working with people in need rather than working to serve them” (Rhoads, 1997, p. 8, emphasis added; Pompa, 2002, p. 4). Everyone involved learns from each other.

Service-learning models are intended to facilitate positive, educative outcomes, although such does not always occur (Dewey, 1938; Simons & Cleary, 2005). Scholarship on best practices for college students indicates that there remain several objective and subjective outcomes of interest, yet some broaden our notions of educative outcomes more than others. The broader, more subjective outcomes studied include course impacts on academic efficacy (Markus, Howard, & King, 1993), levels of a sociological imagination (Marullo, 1998; Scarce, 1997), cultural competence (Starks et al., 2011), cultural relativism as opposed to general feelings of ethnocentrism toward out groups (Borden, 2007), and self-reported levels of social justice and a sense of mutuality (Lewis, 2004), just to name a few.

With one exception (i.e., Allred, 2009), we were unable to find a published evaluation of a service-learning model within prisons. Most likely, this is because service learning is an uncommon pedagogy within the walls, even though there is some variation in forms of correctional education (e.g., onsite with and without the use of specially developed computer software, distance learning; Crayton & Neusteter, 2008). Within the scholarship on correctional education, we highlight a few relevant findings for purposes of context.

First, there remains a significant gap between the aggregate educational level of people incarcerated compared with the general population (BJS, 2003; Wade, 2007). As a group, adults with relatively low educational attainment are overrepresented among people who are incarcerated.

Second, there is a history of empirical and anecdotal evidence that participating in educational programs, postsecondary correctional education (PSCE) in particular (Steurer & Smith, 2003), is correlated with important objective and subjective outcomes. With regard to the former, correctional education is often evaluated in light of the “leading statistical indicator of return on investment” in prisons (Pew Charitable Trust, 2011, p. 6), namely recidivism. Whether recidivism is operationalized as reincarceration, rearrest, reconviction, or some combination, there is ample evidence that it remains the key criteria of interest despite evidence that recidivism is impacted by a variety of individual (e.g., family support, educational attainment, employment) and mesolevel (e.g., postprison supervision periods and other community corrections policies) variables. Consistently, however, multiple studies find that participating in some form of educational programming that involves “higher learning gains” is strongly correlated with lower recidivism (Wade, 2007, p. 29). Others have made stronger claims about the merits of correctional education: an Alabama State Board of Education member said, “Correctional education appears to be the number one factor in reducing recidivism rates nationwide” (as cited in Open Society Institute, 2007, p. 4). Indeed, the justification for programming due to its link with recidivism is such as concern, that if “nothing works” toward this end (Welch, 2004, p. 79), other possible outcomes pale in comparison as indicated by the dearth of published evaluations of correctional educational programming.

This study builds upon an emerging interest with subjective, more psychosocial outcomes associated with PSCE. Fine (2001), Lahm (2009), and Tewksbury and Stengel (2006) represent a significant departure from the research tradition of PSCE described above. For example, they find that PSCE has beneficial impacts for overall prison conditions (e.g., a reduction of disciplinary incidents), as well as the self-evaluations of participating students regardless of their prospects for release in the near or distant future (Banks, 2003). Tewksbury and Stengel (2006) in particular, find that students who take part in PSCE as compared with those who are involved exclusively with vocational programming show significantly higher levels of self-esteem.

Last, there remains a gap in our understanding of how participation in PSCE structures the lived experiences of students in correctional settings and the psychosocial processes that take place as they may involve aspects of the self-concept. Some speculate that enrollment in PSCE reduces recidivism because PSCE increases cognitive skills that change behavior (Bazos & Hausman, 2004). Inferring from research on occupational conditions, educational interventions, or educational evaluations in noncorrectional settings, we argue that achieved self-efficacy in the context of PSCE may be a vital, intervening factor in the overarching relationship between PSCE, recidivism, and other related and equally important human capital (e.g., greater willingness to pursue educational opportunities) or intrapersonal processes (e.g., enhanced self-esteem, heightened feeling of civic engagement). For these reasons, we designed a simple, descriptive assessment of transitions in levels of self-efficacy using data from three prison-based college courses that employ the same pedagogy: an academically challenging, dialogic, collaborative, and transformative learning experience. We describe our basic hypotheses in the next section.

Study Hypotheses

We explore two hypotheses: (a) At Time 1, levels of self-efficacy will be lower among inside students compared with levels among outside students, and; (b) at Time 2, levels of self-efficacy will increase from Time 1 for all students.

With regard to (a), we propose three reasons for this outcome. One, the prison environment is replete by design with conditions and socially structured activities that undermine or constrain efficacy. The physical surroundings, norms, rituals, and elements of prison culture are immediate in the 24/7 prison environment for incarcerated people and in sharp contrast to the academic culture that engulfs typical undergraduate students. Two, net of contributions from the prison culture, any differences observed at Time 1 may be due to individuals’ prior (preincarceration) sense of self-efficacy. Although some inside students have educational levels similar to their outside classmates, as a group, inside students have more diverse educational backgrounds (Pompa & Crabbe, 2004). Also, people with lower levels of educational attainment are over represented among those who are incarcerated (BJS, 2003), and self-efficacy is an established correlate of educational level (Leganger, 2003). Last, anecdotal information from inside–out teachers and staff indicate other reasons why we may expect lower levels of self-efficacy among inside students at Time 1 (Pompa & Crabbe, 2004). Despite a penchant for learning, some inside students come to the firstclass session with concerns about their academic abilities due to a hiatus in their educational experiences due to incarceration, age, or both. For some, it takes the passage of time before they are convinced about and able to earn affirming feedback about their academic capabilities.

With regard to (b), the inside–out pedagogy involves the consistent application of core elements described elsewhere (e.g., Bandura, 1977; Bong, 2004; Jackson, 2002) as social situational variables that influence self-efficacy. Although inside–out courses are not designed intentionally to promote aspects of self-efficacy per se, they are, however, structured to offer ample opportunity for autonomous activity. Some elements that are intricate to the inside–out pedagogy with relevance for self-efficacy are as follows: powerful opportunities for direct personal successes and observing others engaging in accomplishments (i.e., vicarious experiences of success); self-regulatory expectations and benefits that accrue by self-awareness of compliance (e.g., helping create and sustain rules of class engagement); opportunities to display and when necessary improve academic skills via graded course components (e.g., critical analysis of weekly readings, writing assignments that integrate observations, course readings, and personal reflections); and opportunities for students to make cognitive connections across the seemingly diverse tasks and situations (e.g., integrating substantive material effectively in spoken and written exercises).

In sum, we anticipated higher levels of general self-efficacy at Time 2 for all students, because inside–out courses involve these four efficacy-enhancing elements, and, in addition, the inside–out pedagogy represents a different type and context for learning for both inside and outside students.

Last, although we anticipate higher levels of self efficacy for all students at Time 2, we expect more modest gains among outside students. In the daily life of all students, these classes represent a mere 2 hr and 30 min session held once a week. Inside students, come to class, leaving behind if only for a couple of hours, a “total institution,” a place where individuals have few opportunities to express and assess efficacy and have relatively little access to personal autonomy, both of which hold important connotative meaning reflected in the self-attribution process when the context subsumes a significant portion of one’s day-to-day life. However, outside students are already in a situation that affords them a consistent set of opportunities for independent action which, taken together, may culminate in relatively strong self-attributions of competence or self-efficacy. This may occur not because they come to college inherently more self-confident or competent, but because the “structural imperatives” (Gecas & Seff, 1989) of their academic culture make this a greater possibly. Taken together, the course requirements and pedagogy, offer points of contrast and similarity with concurrent learning for outside students.

Method

The Research Setting: Host Institutions

Correctional facilities

Two of the three correctional institutions in this study were men’s prisons, and one was a work release facility. The work release facility is a 96-bed community-based reentry facility for women. It is the first and only all-female reentry facility in that state. However, male inside students attended this course as well. They were bussed to the women’s work release treatment center from a men’s reentry facility located about 5 miles from the site of the course. Another course was held in the same state and county as the work release facility, but in the men’s maximum security prison. This facility holds about 1,500 to 1,700 men on a daily basis, and includes both detentioners awaiting trial as well as sentenced those who are sentenced. Only those already sentenced were eligible for the course. The other course was held in a different state, at a county prison for men. It is a medium security, 448-bed facility. Many people incarcerated at this facility provide labor for the benefit of county public works, recreation authority, animal control, city recycling center, courthouse, airport, and contract details to the city.

Educational Facilities

Two of the courses are affiliated with the same research university, a state-assisted, privately governed land grant, sea grant, and space grant institution. The student body includes about 17,000 undergraduates, 3,700 graduate students, and 850 students in professional and continuing studies. The third course from which data are derived is a private, coeducational liberal arts college in the southeast. The student body includes 1,928 undergraduate and 159 graduate students.

The Research Setting: Instructors and Courses

Participating instructors (n = 3) completed the weeklong inside–out training course that covers the inside–out curriculum, pedagogy, and tenets summarized in the Instructor’s Manual (Pompa & Crabbe, 2004). This manual is a training resource that offers a week-by-week set of ideas for class sessions and exercises as well as a discussion of background concerns and considerations.

This study combines data from three courses, each involving weekly 2 hr and 30 min sessions over 15 weeks. One was an upper level, elective sociology class called Sociology of Disability. It provided an introduction to the topic of disability from a sociological perspective, with a focus on disability issues and policy concerns in correctional facilities (e.g., the aging process during incarceration, the mentally ill in correctional facilities). The second was an upper level criminal justice and sociology course called Drugs and the criminal justice system. It provided an overview of the impact of drugs on individuals and society, the primary criminal justice system, and some promising policy developments. The third was a criminology class, which was also an upper level criminal justice and sociology course. All were elective courses in their respective departments. See Table 1 for a summary of course variables.

Table 1.

Descriptive Information on Courses and Students.

	Course #1	Course #2	Course #3
School type	Research university	Research university	Private, liberal arts college
Correctional facility	Work release facility within a women’s reentry program (men from a nearby reentry facility bussed in to meet here)	Maximum security men’s state prison	Medium security men’s state prison
Course title	Criminology	Drugs and the Criminal justice system	Sociology of disability
Number of students at Time 1	18 inside (10 female & 8 male)	15 inside (all male)	15 inside (all male)
	17 outside (12 female & 5 male)	15 outside (10 female & 5 male)	15 outside (12 female & 3 male)
	35 total	30 total	30 total

All three instructors followed the inside–out model, although the content of discussion and activities varied in accordance with course topics. Each class session involved the application of a “smorgasbord of interactional templates that involve moving and mixing students quickly between icebreakers, small group discussion, large group brainstorming, and so forth” (Allred, 2009, p. 251). In this manner, the structure of class sessions kept students meaningfully engaged with the topic and each other.

Study Sample

At the beginning of the semester, there were 95 people eligible to take part: 48 inside students and 47 outside students. Among inside participants at Time 1, 79% were male and 21% were female. Among outside participants at Time 1, 28% were male and 72% were female. No other demographic information is available on inside and outside students, and it was not possible to link self-efficacy scores with measures of student sex.

Both inside and outside students volunteered for the respective classes. Although the specifics for locating or screening students varied, all three instructors used the same basic criteria for admitting students into the classes. For inside students, admission was a process that involved educational level (most with at least high school diploma or its equivalency), question-and-answer interviews with the instructor (one-on-one or in group format), disciplinary history, level of interest, and staff recommendation. For outside students admission involved completion of academic prerequisites, background check, question-and-answer interview with the instructor (one-on-one or in group format), and affiliation with the department with priority given to majors. All eligible inside and outside students were accepted for the course up to the limits of enrollment.

General Self-Efficacy Survey: The Instrument

This study of self-efficacy involved a survey that measures general levels of self-efficacy. Created by Schwarzer and Jerusalem (1995), the General Self-efficacy Scale (GSE) contains 10 statements (see, Appendix), that are used to create a scale of broad self-efficacy. It is designed to measure people’s belief in their own ability to cope with daily hassles as well as to adapt after experiencing stressful life events. Scale items may be displayed in any order and contain the same response options. The reliability and validity of it have been evaluated elsewhere (http://userpage.fu-berlin.de/~health/self/selfeff_public.htm). Reliability assessments from numerous samples have yielded Cronbach’s α that range from .76 to .90, with the majority in the high .80s. Also, the scale is strong in terms of criterion validity.¹ The original version of the scale used four response options. We use five where 1 is “not true at all” and 5 is “completely true.” Composite scores were tabulated by summing responses to the ten items, with possible scores ranging in value from 10 to 50. The GSE scale was administered in each course at two time periods. Time 1 was the second week of classes (and the first session when inside and outside students met together), and Time 2 was the next to last week of classes.

Ethical Considerations

This study was conducted in accordance with the inside–out Guidelines for Ethical Inquiry (Research Committee of the Inside–Out Prison Exchange Program, 2010).² The Guidelines offer a tool for how to exceed standard human subjects’ considerations and IRB review requirements. We implemented specific procedures that, taken together, responded to two important ethical concerns discussed in the Guidelines, namely confidentiality and voluntary participation. First, our procedures helped distance the instructor from the researcher role. Although the authors openly represented themselves as both teacher and researcher to students, the procedures used here helped prevent negative perceptions or confusion about the teacher role. Second, our survey procedures allowed each student at Time 1 and Time 2 the liberty to “opt in to” or to “opt out of” the study without the instructor or other students able to know participation status.

Results

Participation Rates

At Time 1, data are available for 76 of 95 (80.0% participation) students. At Time 2, data are available for 67 of the 91 (73.62%) students who remained in the class until the end of the semester. Per student group, participation rates range from 71.42% (Time 2, inside students) to 80.85% (Time 1, outside students; see, Table 2). All three participating courses experienced attrition among inside students for a variety of reasons.

Table 2.

Participation Rates Time 1 and Time 2(# completed surveys/number of students in the course = Participation Rates).

Student group	Time 1	Time 2
Inside (I)	38/48 (79.17%)	30/42 (71.42%)
Outside(O)	38/47 (80.85%)	37/49 (75.51%)
Combined (I & O)	76/95 (80.00%)	67/91 (73.62%)

The participation rates of inside student are impacted by both individual decision making and attrition or departure from the course. Inside students are more likely to experience interruptions—positive and negative—to their ability to stay in the course over a 15-week period. Some, as in the case here, experienced early release, transfer to another facility or encountered prohibitions from participating in the course as a form of punishment for violating a facility rule or due to a probation violation. Attrition of this type does not shape the participation rates of outside students.

Statistical Analyses

We used one-sided t-tests to determine whether the comparisons suggested by our hypotheses were due to chance or statistically significant. We applied an alpha level of 0.05 for all t-tests.

Our first hypothesis predicted that inside (M = 34.13, SD = 7.12) and outside (M = 38.37, SD = 5.14) student self-efficacy levels would differ at Time 1, with higher levels observed among outside students. The findings in Table 3 support this prediction, and the differences are statistically significant, t = −2.98, p < .05.

Table 3.

Self-efficacy Mean Scores (10-50, where 50 is high self-efficacy).

Student group	Time 1 (n: SD)*	Time 2 (n: SD)
Inside (I)	(a) 34.13 (38: 7.12)	(d) 39.93 (30: 4.59)
Outside (O)	(b) 38.37 (38: 5.14)	(e) 38.19 (37: 6.12)
Combined (I & O)	(c) 36.25 (76: 6.53)	(f) 38.97 (67: 5.55)
* n = number, SD = standard deviation
Hypothesis #1: a vs. b. t = −2.98, p < .05
Hypothesis #2: c vs. f. t = −2.69, p > .05
a vs. d. t = −4.06, p < .01
b vs. e. t = 0.14, p > .05

Our second hypothesis suggested several comparisons were in order. In general, we predicted that levels of self-efficacy would increase from Time 1 to Time 2, and assessed three possibilities of this outcome. The combined data for inside and outside students indicates increases in self-efficacy by Time 2 (M = 38.97, SD = 5.55), but this difference is not statistically significant from Time 1(M = 36.25, SD = 6.53).

Next, we looked for increases in self-efficacy at Time 2 by evaluating separately the data from inside and outside students. For inside students only, we find a statistically significant increase at Time 2 (M = 39.93, SD = 4.59) in comparison with Time 1(M = 34.13, SD = 7.12). For outside students, we observe virtually no change at Time 2 (M = 38.19, SD = 6.12) in comparison with Time 1 (M = 38.37, SD = 5.14).

These results are remarkable and, to some extent, consistent with expectations gleaned from pertinent research. They also suggest some important processes and possible correlates: the salience of social positioning and malleability of evaluative components of the self in the context of relatively short-term learning experiences, the value of instructor training, indicative of the transformative potential of collaborative, dialogic educational models such as the one examined in this study, and a key bridging mechanism in the relationship between social positioning, individual choices and pursuits, and interpersonal comparison processes informed by relationships with colearners as well as sustained opportunities for the experience of efficacy.

With regard to the remarkable nature of these findings, we are mindful of the total time students are involved with the course, relative to other aspects of their daily lives. At first glance, the educational experience under evaluation would appear to have little potential for influence on self-efficacy: participation involves a 15-week course that meets for one weekly 2 hr and 30 min session. For all student participants, their lives are populated by encounters and possibilities throughout a typical week that may have some bearing on their self-efficacy net of course experiences. Here, we find statistically significant gains in self-efficacy in the student group that was most immersed 24/7 in a daily living environment that, as best we know, is least conducive to efficacy enhancement for reasons already described.

Further, we observe such shifts despite the fact that data are derived from three separate classes, each with a different substantive focus, offered by three different instructors, through two educational institutions and three correctional facilities. Despite their differences, these courses were offered by instructors who took part in the intensive, weeklong inside–out training program. The patterns observed here suggest the importance of instructor training for program consistency, and the program’s ability to teach and model transformative learning, through replicating the learning experience during instructor training.

These findings also offer confidence that the patterns observed among inside students are indicative of two important local, social context influences: (a) the efficacy suppressing elements of a prison culture, and (b) the power of collaborative, prison-based, and college-level educational models to effect intrapersonal change. To be certain, in this evaluation only inside students displayed important shifts in general self-efficacy. If the survey had included measures of self-efficacy in specific domains, the results for inside and outside students may have been more similar and perhaps pronounced. Because the scale did not include items specific to the domain of academic skills (e.g., ability to apply critical thinking skills to course readings, ability to manage class preparation schedule) or areas of knowledge (i.e., theories of offending), we were not able to detect more nuanced possibilities in the shifts in self-efficacy.

Although not captured with the current measures, we believe too that the course model may have influenced outside students, albeit in specific task domains. To begin, we know that student participants—inside and outside—are as likely to experience a variety of potential sources of self-efficacy (e.g., work assignments, classes, friendship networks, family) as they are likely to accrue different types of self-efficacy from similar exposures. All students in inside–out courses participate in a capstone end-of-course project that may culminate in a form of human agency referred to elsewhere as collective efficacy (Sampson, 2004). In our own and other inside–out courses, students take part in an end-of-semester project and work in small groups toward this end. Together, they research, develop, and present a project that addresses some entrenched problem in the criminal justice system. There is compelling anecdotal information that such projects have become a pathway to collective agency: “a central element of the inside–out curriculum, challenging inside and outside students to explore practical ways of implementing changes in what they determine can be improved in the criminal justice system, enlivening discussion about actual proposals for community and social action” (http://www.insideoutcenter.org/local-programs.html). Our findings here concerning the emergence of self-efficacy serve to strengthen the plausibility that some classes may acquire this type of group or collective property.

Last, the findings observed here offer some indication of why it may be important to focus on the process of learning as it is valuable to assess popular program outcomes such as recidivism, continuing education, etc. Across the limited assortment of correctional education models, some correlate more strongly than others with important subjective and objective student outcomes. The findings here encourage us to look more closely to the role of self-evaluation developments and processes as important causal agents. It may be that where learning becomes a platform enabling aspects of “power, authority, and freedom” (Staples et al., 1984, p. 104), the outcomes may manifest in experiences that both embrace and extend well beyond the indicator of recidivism.

Conclusion and Future Study

This study examined transitions in levels of general self-efficacy among students participating in dialogic, college-level courses within a prison setting. All of the courses applied the inside–out prison exchange model. Self-efficacy levels were measured at two time points: at the beginning and end of a 15-week semester. At Time 1, we found lower self-efficacy levels among inside students and provided possible accounts for this result: the prison culture and educational experiences of participating students. At Time 2, we observed a statistically significant increase in self-efficacy among inside students only.

These findings have both theoretical and practical importance. They confirm predictions from social cognitive theory concerning requisite contextual elements for the emergence of human agency in one (personal) of its three forms: personal, collective, and proxy (Bandura, 2000, p. 75). In addition, they provide evidence of subjective course outcomes that may have direct implications for personal life trajectories.

The results reported here should be interpreted in the context of study limitations. First, the survey instrument did not contain demographic questions, nor did it allow the assessment of paired pre- and postcomparisons of self-efficacy. We believe it is important to have information on student identity characteristics, in part, because they have documented (e.g., Stevens et al., 2004) impacts on academic self-efficacy.

Second, the survey asked about general self-efficacy. For this reason, we do not know exactly what students felt efficacious about in their personal sense of confidence. Others find that feelings of self-efficacy in one domain (e.g., being able to converse effectively with others about academic topics, being able to conceptualize action orientations concerning a particular social issue) are not inherently correlated with self-efficacy in other domains (Bong, 2004). Indeed, Bong (2004) says “most academic motivational constructs are known to contain strong domain specific components” (p. 289). Thus, we advise future study to add measures of academic self-efficacy (skill sets and substantive topics) in order to explore the possibility of more pronounced and widespread experiences of self-efficacy associated with this type of learning experience.

Three, it is important to measure the structure and content of a learning context that may impact student experiences (Allred, 2009). The context for this evaluation is a unique pedagogical model that may represent a platform for distinct student experiences and outcomes with relevance for some, but not all models of learning. That is, our pedagogical context and the processes documented herein may not be relevant for inquiry associated with other models of PSCE.

Lastly, all aspects of this study were conceived and implemented by three instructor-researchers who share a solid background (e.g., courses taught range from 4 to 12 each) with the inside–out model. In addition, these instructors-researchers share an intuitive understanding of how “this learning changes lives,” in part, because they (we) have experienced inside–out as trainees and instructors and through numerous, meaningful conversations with students. Despite this epistemological grounding, we recommend that future study of self-efficacy in correctional settings, and inquiry about inside–out in general, involve Participant Action Research (PAR; Reason & Bradbury, 2008) to the extent possible. While we incorporated PAR elements in the developmental phase of this inquiry,³ we were unable to obtain written permission to acknowledge these students by name. Their unanticipated transfers removed the chance to obtain the necessary written consent.

At minimum, a PAR approach to inquiry about human agency would involve collaboration among a broader set of people and include cycles of reflection on procedures, processes, and outcomes with the engagement among such people. For example, a PAR approach may include on an ongoing basis both researchers (teachers) and participants as collaborators (inside and outside students). Such an approach may echo more directly the tenets associated with the inside–out pedagogy, in part, because PAR methodologies are intentionally pursued so as to diminish power differentials, privilege, and truncated perspectives.

To conclude, these study findings offer an important account of why some PSCE programs are particularly impactful: they provide an efficacy-enhancing context within the walls. Experienced as such, this form of human agency may have a direct or indirect influence on subjective and objective positive outcomes for participating individuals as well as host facilities and communities. We hope that future research will refine the findings reported here, and integrate our insights about substantive and methodological considerations.

Footnotes

Appendix

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

Author Biographies

Sarah L. Allred is associate professor, Department of Sociology and Anthropology at Berry College, GA. The social construction of mental illness, health and health care issues, social inequality, and research methodology are her focus, and she’s published in the Journal of Correctional Education, Disability & Society, Ecopsychology, and the International Journal of Environmental Health Research.

Lana D. Harrison is associate professor, University of Delaware–Newark Department of Sociology and Criminal Justice. As the Associate Director of the Center for Drug and Alcohol Studies, her research focuses on drug epidemiology, treatment alternatives, improving survey methodology, comparative international research, and drug policy. She has authored or coauthored over 60 publications in these areas.

Daniel J. O’Connell is associate scientist with the University of Delaware–Newark Center for Drug and Alcohol Studies. Currently Project Director of the NIDA-funded CJ–DATS Project, his research specialties include design and methodology, intervention, and project management. His published articles are on drug treatment, prison management, HIV prevention interventions, and criminological theory.

References

Allred

S. L.

(2009). The Inside-Out Prison Exchange Program: The impact of structure, content, and readings. Journal of Correctional Education, 60, 240-258.

Bandura

(1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215.

Bandura

(1986). Social foundations of thought and action: A social-cognitive view. Englewoods Cliffs, NJ: Prentice Hall.

Bandura

(1990). Reflections on non-ability determinants of competence. In Sternberg

Kolligian

Jr. (Eds.), Competence considered: Perceptions of competence and incompetence across the lifespan (pp. 315-362). New Haven, CT: Yale University Press.

Bandura

(2000). Exercise of human agency through collective efficacy. Current Directions in Psychological Science, 9(1), 75-78.

Banks

(2003). Learning under lockdown. Retrieved from http://www.colorlines.com.

Bazos

Hausman

(2004). Correctional education as a crime control program. Los Angeles, CA: UCLA, School of Public Policy and Social Research, Department of Policy Studies.

Bong

(2004). Academic motivation in self-efficacy, task value, achievement goal orientations, and attributional beliefs. Journal of Educational Research, 97, 287-297.

Bureau of Justice Statistics (BJS; 2003). Education and correctional populations. (U.S. Department of Justice, NCJ 95670). pp 1-5. Washington, DC: Author.

10.

Crayton

Neusteter

S. R.

(2008). The current state of correctional education. Paper presented at the Reentry Roundtable on Education, John Jay College of Criminal Justice, New York, NY.

11.

Dewey

(1938). Experience and education. New York, NY: Collier.

12.

Fine

(2001). Changing minds: The impact of college in a maximum security prison. New York, NY: Ronald Ridgeway.

13.

Gecas

Seff

M. A.

(1989). Social class, occupational conditions, and self-esteem. Sociological Perspectives, 32, 353-364.

14.

Gist

Mitchell

T. R.

(1992). Self-efficacy: A theoretical analysis of its determinants and malleability. Academy of Management Review, 17, 183-211.

15.

Grabowski

L. J.

Call

K. T.

Mortimer

J. T.

(2001). Global and economic self-efficacy in the educational attainment process. Social Psychology Quarterly, 64, 164-179.

16.

Jackson

J. W.

(2002). Enhancing self-efficacy and learning performance. Journal of Experimental Education, 70, 243-254.

17.

Kaposy

Bandrauk

(2012). Prioritizing vaccine access for vulnerable but stigmatized groups. Public Health Ethics, 5, 263-282. doi:10.1093/phe/phs010

18.

Kuhlman

Ruddell

(2005). Elderly jail inmates: Problems, prevalence, and public health. California Journal of Health Promotion, 3, 49-60.

19.

Lahm

K. F.

(2009). Educational participation and inmate misconduct. Journal of Offender Rehabilitation, 48, 37-52.

20.

Leganger

(2003). Control constructs: Do they mediate the relation between educational attainment and behavior? Journal of Health Psychology, 8, 361-372.

21.

Lewis

T. L.

(2004). Service learning for social change? Lessons from a liberal arts college. Teaching Sociology, 32, 94-108.

22.

Luszczynska

Scholz

Schwarzer

(2005). The general self-efficacy scale: Multicultural validation studies. Journal of Psychology, 139, 439-457.

23.

Markus

G. G.

Howard

J. P. F.

King

D. C.

(1993). Integrating community service and classroom instruction enhances learning: Results from an experiment. Educational Evaluation and Policy Analysis, 15, 410-419.

24.

Marullo

(1998). Bringing home diversity: A service-learning approach to teaching race and ethnic relations. Teaching Sociology, 26, 259-275.

25.

Mortimer

J. T.

Lorence

(1979). Occupational experience and the self-concept: A longitudinal study. Social Psychology Quarterly, 42, 307-332.

26.

The National Service-Learning Clearinghouse. (2007). Service learning is. . . Retrieved from http://www.servicelearning.org/what_is_service-learning/service-learning_is/index.php.

27.

Open Society Institute (OSI,. 1997). Education as crime prevention: Providing education to prisoners (Research Brief No. 2) pp 1-15. New York, NY: Author.

28.

Pajares

Miller

M. D.

(1995). Mathematics self-efficacy and mathematics performances: The need for specificity of assessment. Journal of Counseling Psychology, 42, 190-198.

29.

Pew Charitable Trust. (2011). State of recidivism: The revolving door of America’s prisons. Philadelphia, PA: Author.

30.

Pompa

(2002). Service learning as a crucible. Michigan Journal of Community Service, 9, 67-76.

31.

Pompa

Crabbe

(2004). The Inside-Out Prison Exchange Program: Exploring issues of crime and justice behind the walls, instructor’s manual (Rev. ed.). Philadelphia, PA: Temple University.

32.

Reason

Bradbury

(2008). The SAGE handbook of action research: participative inquiry and practice. London, UK: SAGE.

33.

Research Committee of the Inside-Out Prison Exchange Program. (2010). Guidelines for ethical inquiry, Philadelphia, PA: Temple University.

34.

Rhoads

R. A.

(1997). Community service and higher learning: Explorations of the caring self. Albany: State University of New York.

35.

Rosenberg

Pearlin

L. I.

(1978). Social class and self-esteem among children and adults. American Journal of Sociology, 84, 53-77.

36.

Sampson

R. J.

Raudenbush

S. W.

Earls

(1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science, 277, 918-924.

37.

Sampson

R. J.

(2004). Neighborhood and community: Collective efficacy and community safety. New Economy, 11, 106-113.

38.

Scarce

(1997). Field trips as short-term experiential education. Teaching Sociology, 25, 219-226.

39.

Schnittker

John

(2007). Enduring stigma: The long-term effects of incarceration on health. Journal of Health and Social Behavior, 48, 115-130.

40.

Schwarzer

Jerusalem

(1995). Generalized self-efficacy scale. In Weinman

Wright

Johnston

, (Eds.), Measures in health psychology: A user’s portfolio. Causal and control beliefs (pp. 35-37).Windsor, England: NFER–NELSON.

41.

Simons

Cleary

(2005). Student and community perceptions of the “Value Added” for Service-Learners. Journal of Experiential Education, 28, 164-188.

42.

Staples

C. L.

Schwalbe

M. L.

Gecas

(1984). Social class, occupational conditions, and efficacy-based self-esteem. Sociological Perspectives, 27, 85-109.

43.

Starks

B. C.

Harrison

Denhardt

(2011). Outside the comfort zone of the classroom. Journal of Criminal Justice Education, 22, 203-225. doi:10.1080/10511253.2010.517773

44.

Steurer

S. J.

Smith

L. G.

(2003). Education reduces crime: Three-state recidivism study (Executive Summary) Lanaham, MD: Correctional Education Association.

45.

Stevens

Olivarez

Jr. Lan

W. Y.

Tallent-Runnels

M. K.

(2004). Role of mathematics self-efficacy and motivation in mathematics performance across ethnicity. Journal of Educational Research, 97, 208-221.

46.

Tewksbury

Stengel

K. M.

(2006). Assessing correctional education programs: The students’ perspective. Journal of Correctional Education, 57, 13-25.

47.

Vega

Silverman

(1988). Stress and the elderly convict. International Journal of Offender Therapy and Comparative Criminology, 32, 153-162.

48.

Wade

(2007). Studies of correctional education programs. Adult Basic Education and Literacy Journal, 1, 27-31.

49.

Welch

(2004). Corrections: A critical approach (2nd ed.), New York, NY: McGraw Hill.

50.

World Health Organization. ( Møller

Stöver

Jürgens

Gatherer

Nikogosian.

[Eds,]; 2007). Health in prisons: A WHO guide to the essentials of prison health. Copenhagen, Denmark: World Health Organization Regional Office for Europe.

51.

Zajacova

Lynch

S. M.

Espenshade

T. J.

(2005). Self-efficacy, stress, and academic success in college. Research in Higher Education, 46, 677-706.