INTUIT: A Career Decision-Making Intervention for Female Offenders in Reentry

Abstract

This study evaluates cross-sectional data from 9 years of the INTUIT: Work and Careers intervention (N = 491), a program for female offenders in reentry. Of several measures used, participants showed increases on the five career decision-making self-efficacy factors assessed by the Career Decision Making Self-Efficacy Short Form (CDMSE-SF), over the course of the intervention, whereas no changes were observed across the same testing period for the control group. When control and invention groups were compared, statistically significant differences were only seen for INTUIT participants in levels of self-appraisal and goal selection. Implications of these findings are addressed.

Keywords

female offenders reentry career decision-making career-related self-efficacy

Examining recidivism in the United States between 2005 and 2010, Durose, Cooper, and Snyder (2014) reported that approximately 58.5% of female offenders were rearrested within 3 years, and 68.1% within 5 years, of their release from prison. Although achieving steady employment consistently correlates with reduced recidivism in female participants (e.g., Greenfield & Snell, 1999), most women preparing for societal reentry have limited access to quality career-related programming (Hurry, Brazier, Parker, & Wilson, 2006). Important progress has been made in response to calls for evidence-based vocational programming for offenders; and counselors, psychologists, and criminal justice professionals all value the contribution that researchers and practitioners offer reentering offenders (cf. Varghese, Fitzgerald, Chronister, Cummings, & Forrest, 2013). However, mixed results and spotty delivery of interventions underscore an ongoing need for evidence-based career-related interventions (Lawrence, Mears, Dubin, & Travis, 2002; Petersilia, 2004; Varghese, 2013; Visher & Kachnowski, 2007).

We review two career-related counseling interventions for offenders in reentry, both of which use the social cognitive career theory (SCCT) framework (Lent, Brown, & Hackett, 1995, 2002). First, Chartrand and Rose (1996) developed Project PROVE, a 12-week intervention aimed at clarifying occupational interests of female offenders while simultaneously acknowledging environmental barriers faced by the prison population: limited education, poverty, discrimination, and subculture socialization that may separate offenders from the mainstream population. The goals of Project PROVE were to address and enhance female offenders’ job search and problem-solving skills, to increase their knowledge of work, and improve their career-related self-efficacy (Chartrand & Rose, 1996). According to available documentation, the program focused a positive approach to goal setting, rational decision-making, and counteracting some of the impulsive tendencies identified as characteristic of the offender population (Ross & Fabiano, 1985). Results of this intervention gleaned through reviews of technical reports further suggest that female offenders increased their job search self-efficacy and outcome expectations, and reported increases in problem-solving and career planning. Finally, and relative to the comparison group, women who participated in Project PROVE reported greater involvement in career-related activities.

More than 15 years later, Fitzgerald, Chronister, Forrest, and Brown (2012) made another contribution to the counseling literature in reporting on their OPTIONS program. This program also was developed using the SCCT framework and focused on vocational exploration, clarification of interests, and development of knowledge pertaining to job search skills––including career options, career planning, and use of supports. The program showed success in that participants were more hopeful, had increased career search self-efficacy, and improved problem-solving relative to controls. However, the OPTIONS intervention was designed for male offenders and has not been adapted to meet the unique needs of female offenders (e.g., Vigesaa, Bergseth, & Jens, 2016). The OPTIONS program also was limited in scope and duration. It was delivered in five 2-hr sessions.

In this paper, we report on the 13-week INTUIT program, a career-related intervention for female offenders in reentry, and successor to Project PROVE. The similarities among these programs are more significant than their differences insofar as each aimed at improving offenders’ knowledge of self and work-related knowledge regarding employment options upon release. Both PROVE and INTUIT focused on the career-related needs of female offenders in reentry. Therefore, the purpose of this paper is 3-fold. We first describe the core components of INTUIT; second, the findings regarding career decision-making self-efficacy, a measure of intervention effectiveness in evaluation of INTUIT; and third, the limitations of the INTUIT program and recommendations for future work.

Theoretical and Empirical Bases of INTUIT

As with PROVE and OPTIONS, INTUIT was constructed with an SCCT framework, addressing the contextual proximal influences of barriers and social supports on occupational interests and goals (Lent et al., 1995, 2002). The SCCT framework is fitting for employment-related interventions, and especially for female offenders, because of their limited resources and social supports. Earlier career development theories emphasized the relation between individual differences and a person’s environment, but largely neglected the effects of poverty, racial discrimination, and exposure to chronic stress (i.e., career barriers). In SCCT, contextual or environmental factors (terms used interchangeably within this paper) such as those mentioned above have the potential to become internalized impediments to the reentering offender’s career progress. Conceptually, these factors influence career choice making (Lent, Brown, & Hackett, 2000).

Integral in this study is the critical concept of self-efficacy––an important construct within the SCCT framework which addresses the perceptions an individual has about their ability to perform a learning task (Bandura, 1986). In SCCT, decision-making self-efficacy is influenced by individuals’ expectations of the employment opportunities available to them. Offenders reentering the workforce after incarceration are beset with barriers to employment, most importantly overcoming employers’ perceptions of hiring persons with a criminal record. Lent, Brown, and Hackett (1995, 2002) suggest that if an individual perceives fewer or less desirable opportunities, the effect of their self-efficacy beliefs and outcome expectations have a more direct effect than their interests or goals on their decisions regarding an occupation. To this effect, the authors of this paper evaluated the effectiveness of the INTUIT intervention by targeting female offenders’ career decision-making self-efficacy. Furthermore, SCCT is a theory of career choice; therefore, interventions that evaluate decision-making self-efficacy and decision-making approaches may be particularly helpful in advancing research with this population.

Brown and Ryan-Krane’s (2000) meta-analytic work on career interventions suggested that five components seem especially predictive of positive effects related to career decision-making. Hence, the following features were explicitly incorporated into the INTUIT program: (a) allowing individuals to clarify their career goals and plans by way of reflective exercises; (b) providing individualized interpretations of career-related assessments; (c) affording accurate and up-to-date information regarding the world of work; (d) offering opportunities for individuals to interact with successful models––that is, individuals who have used effective strategies to reach their goals; and (e) helping individuals to assess their existing social network, to strengthen it, or to develop a new and healthy social network. The following paragraphs address the components of the INTUIT program in respect to these features.

INTUIT Program Overview

Reflective/written exercises––Writing exercises recognizing offenders’ level of comfort with verbal expression were provided in the program. Tasks required only brief responses, for example, having offenders complete a listing of career-related barriers they perceive. More involved reflective tasks incorporated visual components, for example, a time-line task using visual stimuli to gather participants’ personal history, positive and negative life events, and plans. During group sessions, as well, participants listed out perceived occupational barriers, writing them on a paper that had the pattern of a brick wall.

Individualized Interpretation and Feedback––Offenders were given published assessment instruments to complete. Specifically, versions of the Self-Directed Search and The NEO Five Factor Inventory were administered. Data were interpreted and provided in group format, as well as through optional private feedback sessions. Decision-making styles, including the advantages of a more systematic/rational decision-making style (termed “vigilance”) and the disadvantages associated with a passive stance that shifts responsibilities (termed “buck-passing”), were discussed and measured at two time points (Time 1 and Time 2) as an intervention outcome variable using participant responses to the Melbourne Decision Making Questionnaire (MDMQ; Mann, Burnett, Radford, & Ford, 1997). The Tendency to Foreclose subscale (TTFS) of the Commitment to Career Choices Scale (CCCS; Blustein, Ellis, & Devenis, 1989), and Betz, Klein, and Taylor’s (1996) short form of the Career Decision-Making Self-Efficacy Scale (CDMSE) were used to explore tolerance of ambiguity and openness to diverse possibilities in the career commitment process, as well as the constructs of occupational information-seeking, career goal selection, career planning, career problem-solving, and career self-appraisal. We believed a tendency to foreclose on occupational choices could be particularly relevant for female offenders, who may default into work options having had little opportunity for exploration (Blustein et al., 1989). In addition, and given that negative career cognitions can diminish decision-making self-efficacy (Andrews, Bullock-Yowell, Dahlen, & Nicholson, 2014), career decision-making self-efficacy was thought to be especially salient for female offenders facing challenges in obtaining employment.

Up-to-date information regarding the world of work––We provided substantial amounts of practical, work-related information to participants. It is important to note that most employment information is now available only via electronic media (e.g., ONet Online, LinkedIn.com* ). This complicates matters for inmates, as corrections agencies, as a rule, do not allow offenders access to the internet. We showed offenders how to access and use ONet Online* and other resources including online classified advertisements by way of in-class demonstrations. We also discussed ways to find free Internet access upon release (e.g., at public libraries, coffee shops, government services agencies).

Opportunities to interact with successful models––Graduate students delivering the INTUIT program, with the assistance of undergraduate students, served as one source of modeling for program participants. The community corrections sites at which the intervention was delivered also offered access to successful program graduates (inmates who successfully reentered society) who returned to discuss their experiences. These visits, and participants’ reactions, were carefully processed during the intervention. As in other areas of life, one of the most powerful and effective ways to teach and inspire individuals was to allow them to interact with successful role models (Bandura, 1986).

Attention to building support networks––We also emphasized development of healthy support networks to replace the subculture socialization that can separate offenders from the mainstream population (Chartrand & Rose, 1996). Individuals’ attempts to keep jobs and desist from illegal activities can be related to the quality and amount of support received from important others (Shivy et al., 2007). Related group activities helped offenders examine, in formal ways, the status of their postrelease social network and available sources of structure and support. For example, we discussed challenges that participants face in returning to social groups involved with crime or substance use disorders. This issue seemed particularly salient for women whose partners were substance involved.

Besides attending to the five critical components of career-related interventions, INTUIT participants focused on how to communicate efficacy beliefs, which are their career-related interests and abilities to potential employers––activities consistent with the tenets of SCCT (Lent et al., 1995, 2002). Participants completed job applications and role-played responses to difficult interview questions and on-the-job scenarios, including anticipated sex discrimination. It is interesting to note that, although participants uniformly expressed confidence in regard to completing job applications, their actual performance turned up problems. Gentle and supportive critique helped participants fix critical errors and reinforce the need for practice. A key aspect of the intervention was having participants rehearse when and how to inform employers of their criminal histories (Shivy et al., 2007), and one full INTUIT session addressed the concept of stigma––both from society and from self.

Specific Aims and Hypotheses

As was the case with PROVE, the efficacy of INTUIT was to be assessed by comparing participant outcomes at two time points (Time 1 and Time 2). The aim of this study was to determine whether, and to what degree, exposure to the INTUIT career intervention would result in changes on psychological variables related to career decision-making. The following hypotheses were proposed:
Hypothesis 1: Women participating in INTUIT program will demonstrate a lower tendency to foreclose on career options after the program than before the program, indicating more openness to considering career alternatives in their decision-making.

Hypothesis 2: Women participating in the INTUIT group will demonstrate lower affinity for the “buck-passing” style of decision-making and higher identification with the “vigilant” style of decision-making on the MDMQ (Mann et al., 1997).

Hypothesis 3: INTUIT participants will have higher postintervention scores on the five subscales on the short form of the CDMSE (Betz, Klein, & Taylor, 1996) in comparison with the control group, indicating that their confidence for career decision-making and job-seeking activities has increased over the course of the program.

Hypotheses resulting from comparisons from Time 1 to Time 2 among within-group participants are detailed below.
Hypothesis 4: Relative to the control group, INTUIT participants will demonstrate a lower tendency to foreclose from Time 1 to Time 2.

Hypothesis 5: Relative to the control group, INTUIT participants will show lower scores from Time 1 to Time 2 of identification with the “buck-passing” style of decision-making.

Hypothesis 6: Relative to the control group, INTUIT participants will show higher scores from Time 1 to Time 2 of “vigilant” decision-making and career decision-making self-efficacy.

Method

Participants

University and Department of Corrections’ Institutional Review Board approval were obtained and participation in this project was voluntary for all inmate participants. Four hundred and ninety-one nonviolent female felony offenders under community corrections residential supervision in a Southeastern state participated. All were beginning their stay in a 20-week step-down program. Women were admitted to the community corrections facility in small groups, termed “waves,” upon transfer or release from other entities within the correctional system. Other forms of corrections programming provided to the women included substance abuse education, Narcotics Anonymous and Alcoholics Anonymous groups, General Education Diploma (GED) preparation, parenting skills, independent living skills, transitional services, and coping with domestic violence. During the program, which included work placement during the day, court costs and restitution are collected and community service duties are performed.

It is important to note that assignment to conditions was not under direct control of the investigator and is generally not permissible in this Southeastern state’s Department of Corrections (DOC). Therefore, further reference to a no-treatment control condition should be interpreted as being a quasi-control group. Individuals were allocated to groups based on DOC release or transfer schedules within the community corrections system––a relatively random event dictated by court processing time and reception and diagnostic services. An education coordinator at the facility assigned women to the intervention group or the control group, in waves, based on the number of admissions to the facility. The group size included approximately 15 participants. Table 1 shows demographic characteristics for the intervention and control groups. The groups did not differ significantly on mean years of age, race, average number of children, or educational attainment; however, they differed in terms of marital status (p < .05). The intervention group had more married and partnered women and fewer single women.

Table 1.
Demographics of the Intervention and Control Groups.

Intervention Control

Race

African American 36.4% 37.8%

Asian American/Pacific Islander 0.4% 0.0%

Caucasian 59.7% 58.4%

Hispanic 1.6% 0.8%

Middle Eastern 0.4% 0.0%

Other 1.6% 2.9%

Age in years at start of program 31.34 (7.91) 30.42 (8.07)

Marital status

Single, never married or partnered 22.2% 36.1%

Married or partnered 48.6% 36.6%

Separated 13.4% 13.0%

Divorced 14.2% 12.6%

Widowed 1.6% 1.7%

Average number of children 1.84 (1.62) 1.68 (1.40)

Highest educational achievement

Some high school 31.6% 31.9%

High school diploma/GED 36.4% 33.6%

Some college/2-year degree 16.6% 23.1%

College graduate 4.3% 6.7%

Other (indicated graduate degrees) 9.1% 3.8%

Measures

Tendency to foreclose

The TTFS of Blustein, Ellis, and Devenis’s (1989) CCCS was used to identify the extent to which an individual foreclosed on a career path without engaging in patterns of exploration and choice. Items include “I believe that there is only one specific career goal that is right for me,” and “I like the openness of considering various possibilities before committing myself to a specific occupation.” The TTFS is a 9-item scale that has adequate reliability and validity as a stand-alone scale and has been found useful in establishing relationships with less adaptive educational and career-related behaviors (Lopez, 1997). In the present study, the Cronbach’s α of the TTFS was .71 at Time 1 and .73 at Time 2.

Decision-making style

In 1997, Mann, Burnett, Radford, and Ford published the MDMQ, a 22-item instrument identifying four decision-making styles: vigilance, hypervigilance, buck-passing, and procrastination (both forms of defensive avoidance). We administered only the vigilance (α = .80) and the buck-passing (α = .87) scales with participants, as we were interested in exploring participants’ use of rational or dependent decision-making styles (cf. Mau, 1995). Prior to administration of measures, it was determined that the Hypervigilance and Procrastination scales of the MDMQ would not be appropriate for offenders within 6 months of societal reentry. The Hypervigilance Scale evaluates a hurried or time-pressured approach to decision-making, whereas the Procrastination Scale assesses delays in decision-making. Both decision-making approaches are contingent upon the influence of true external time pressures. Since offenders decidedly maintain little control over the time in which they reenter society and are, thus, limited regarding the influence of such pressures, administering scales of temporally based approaches to decision-making did not seem appropriate for use in this study.

Vigilance is conceptualized as a decisional style that entails rational decision-making (Mann et al., 1997) and is associated with positive self-evaluation (Di Fabio & Palazzeschi, 2012). Although several decision-making scales are available in the literature, the MDMQ has advantages that promoted its use with offenders. The MDMQ items are written simply and clearly using vocabulary at a middle school reading level, and evidence of validity is presented for individuals of varying ethnicities and ages. The MDMQ also refers to general decision-making skills rather than career-specific skills, which fits with some offenders’ limited work experience. In the present study, observed Cronbach’s αs were .73 for the Buck-passing Scale (MDMQ Buck-Passing) and .69 for the Vigilance Scale (MDMQ Vigilance) at Time 1; and .94 and .88 at Time 2, respectively.

Career decision-making self-efficacy

Betz et al. (1996) devised and evaluated a short form of the CDMSE, which measures the extent to which a person believes that he or she is capable of making good career decisions and successfully executing career-related behaviors. The short version of the scale retains the five-factor structure, assessing occupational information-seeking, goal selection, planning, problem-solving, and self-appraisal while reducing the number of items for each subscale from 10 to 5. The 25-item CDMSE-SF was deemed appropriate for the current sample. It is easy to understand, relatively brief, and has been validated with specialized groups, including battered women, who bear psychological resemblance to the female offender population (Brown, Reedy, Fountain, Johnson, & Dichiser, 2000). For this population, the 10-point scale normally used for the CDMSE-SF was shortened to a 5-point scale for ease of use with the population and to maintain continuity among the administered scales. In the present study, observed Cronbach’s αs were .65 for self-appraisal, .66 for information, .77 for goal selection, .83 for planning, .77 for problem-solving, and .93 overall at Time 1; and .84, .87, .78, .85, .59, and .94, respectively, at Time 2.

Results

Preliminary Analyses

Means and standard deviations on continuous demographic variables and outcome variables at Time 1 were compared between intervention participants and the control participants. The distribution of the categorical demographic variables was compared between the two groups as well. Given the number of comparisons conducted, the Type I error rate of each individual test was set at .01 level to control family-wise error rate.

Independent samples t tests were conducted to compare age, number of children, and all outcome variables between the control group and the intervention group at Time 1. No significant differences were detected between the groups based on age, number of children TTFS, decision-making styles (MDMQ Buck-Passing or MDMQ Vigilance), or the career decision-making self-efficacy subscales, including self-appraisal (S-APP), occupational information (INFO), goal selection (GOAL), planning (PLANS), or problems-solving (P-SOLV). Table 2 provides descriptive statistics for both groups.

Table 2.
Descriptive Statistics and Independent Samples t tests of Groups at Time 1.

Control Group Intervention Group

N Mean SD N Mean SD p

Age 238 30.42 8.07 253 31.34 7.91 NS

Children 238 1.68 1.40 253 1.84 1.62 NS

TTFS 237 32.23 9.32 245 31.76 9.28 NS

MDMQ Buck Passing 238 14.87 2.68 252 14.36 2.75 NS

MDMQ Vigilance 238 8.76 2.31 251 8.94 2.31 NS

S-APP 238 19.48 3.72 253 18.76 4.94 NS

INFO 237 19.36 4.04 253 18.86 5.57 NS

GOAL 238 19.03 3.82 252 18.27 4.15 NS

PLANS 238 18.44 4.08 253 17.87 4.44 NS

P-SOLV 238 17.95 4.04 253 17.92 4.18 NS

Note. NS = not significant; TTFS = Tendency to Foreclose Subscale; MDMQ = Melbourne Decision Making Questionnaire; S-APP = self-appraisal; INFO = occupational information; GOAL = goals selection; PLANS = planning; P-SOLV = problem-solving.

Chi-square analyses were used to compare the control and the intervention groups on group composition of ethnicity, marital status, highest level of education, and their identified educational goals at Time 1. No significant differences were detected between the control and the intervention groups for ethnicity, highest level of education, and identified educational goals. Marital status differed significantly between the control and intervention groups. The intervention group had more married and partnered women and fewer single women.

Hypothesis Testing

Following preliminary analyses, tests of hypotheses were conducted assessing within-group changes for the intervention and control groups from Time 1 to Time 2. Comparisons of outcome variable scores among those in the control group were conducted for the purpose of determining whether simple maturation (i.e., the time to release) could account for any changes among those not participating in the intervention. To test Hypothesis 3, determining whether INTUIT participants’ involvement in the intervention improved self-efficacy in comparison with the control group, an analysis of covariance (ANCOVA) was conducted using self-efficacy scores from both groups at Time 2. The results of hypotheses testing are provided in the following paragraphs.

Time 1 Versus Time 2 Comparisons on the Outcome Variables

Intervention group

Table 3 shows Time 1 and Time 2 descriptive statistics of outcome variables in the intervention group. Paired samples t tests were used to detect intervention effects on the outcome variables. Differences were detected at .01 significance level for five of the five career decision-making self-efficacy factors, including self-appraisal (t = 5.22, p < .001, η² = .10, Cohen’s d = .33), occupational information (t = 3.29, p = .001, η² = .04, Cohen’s d = .23), goals (t = 4.75, p < .001, η² = .09, Cohen’s d = .30), planning (t = 5.44, p < .001, η²= .11, Cohen’s d = .34), and problem-solving (t = 2.67, p = .008, η² = .03, Cohen’s d = .18). All differences were in the expected directions. A difference also was detected at .01 significance level for the MDMQ Vigilance Scale (t = 4.22, p < .001, η² = .07, Cohen’s d = .33). Surprisingly, results suggest that individuals in the INTUIT group showed changes to a less vigilant decision-making style over time.

Table 3.
Descriptive Statistics and Paired Samples t Tests for the Intervention Group.

Time 1 Time 2

N Mean SD Mean SD p

TTFS 229 31.61 9.19 30.55 8.55 NS

MDMQ Buck-Passing 252 14.36 2.75 14.02 4.11 NS

MDMQ Vigilance 251 8.94 2.31 8.09 2.83 <.001

S-APP 243 18.59 4.36 19.95 3.87 <.001

INFO 242 18.82 5.66 19.97 4.20 .001

GOAL 242 18.26 4.20 19.44 3.72 <.001

PLANS 243 17.86 4.54 19.31 3.96 <.001

P-SOLV 243 17.94 4.19 18.67 3.77 .008

Note. NS = not significant; TTFS = Tendency to Foreclose Subscale; MDMQ = Melbourne Decision Making Questionnaire; S-APP = self-appraisal; INFO = occupational information; GOAL = goals selection; PLANS = planning; P-SOLV = problem-solving.

Control group

Similarly, means and standard deviations of all outcome variables at Time 1 and Time 2 were examined for the control group, although these values are not tabled. Again, paired samples t tests were used to examine any changes. In contrast to the intervention group, differences were detected at .01 significance level only for the decision-making scales, MDMQ Buck-Passing (t = 4.68, p < .001, η² = .08, Cohen’s d = .39), and MDMQ Vigilance (t = 3.98, p < .001, η² = .06, Cohen’s d = .32). Results indicate that individuals in the control group showed changes to both less dependent and less vigilant decision-making styles over time.

Control group versus intervention group comparison of self-efficacy scores

A one-way independent group ANCOVA was used to compare the control and the intervention groups for the self-efficacy scales at Time 2. In the ANCOVA procedure, Time 2 scores of the measures were entered as dependent variables. The Time 1 scores were entered as covariates, while the dichotomous variable of group membership (control vs. intervention) served as the independent variable. Porter and Raudenbush (1987) suggest using a covariate linearly related to the dependent variable in order to control for the effects of the intervention. An assumption of ANCOVA includes independence of the covariate and the intervention effects. Time 1 scores were selected as a covariate because they were presumably unaffected by the intervention. The homogeneity of regression assumption was tested by including the interaction term between the covariate and the between-subjects factor (control and intervention group). Interaction terms in all analyses were not significant, indicating that the homogeneity of regression assumption was not violated. Analyses indicated that there were significant differences between the control and the intervention groups in terms of their levels of self-appraisal, F(1, 446) = 8.21, p = .004; and goal selection, F(1, 445) = 10.22, p = .001, when controlling for the effects of the Time 1 scores (see Tables 4 and 5). However, analyses also indicated that there were no significant differences between the control and intervention groups for occupational information, F(1, 444) = 5.00, p < .01, planning, F(1, 446) = 6.77, p < .01, or problem-solving, F(1, 446) = .48, p < .01, with the effects of the Time 1 scores removed. Therefore, the control and intervention groups differed on two of the five self-efficacy scales: self-appraisal and goal selection.

Table 4.
Descriptive Statistics for Control and Intervention Groups at Time 2 and ANCOVA Results for Self-Appraisal (S-APP).

Control Group Intervention Group

N Mean SD N Mean SD F p

S-APP2 206 19.50 4.01 243 19.95 3.87 8.21 .004

S-APP1 206 19.48 3.72 243 18.76 4.94 131.03 < .001

Table 5.
Descriptive Statistics for Control and Intervention Groups at Time 2 and ANCOVA Results for Goal Selection (GOAL).

Control Group Intervention Group

N Mean SD N Mean SD F p

GOAL2 206 18.91 3.90 242 19.44 3.72 10.22 .001

GOAL1 206 19.03 3.82 242 18.27 4.15 155.06 < .001

Discussion

This cross-sectional evaluation of the major career-related outcomes of the INTUIT intervention, across 9 years of data, suggests that participants benefited from participation in programming. INTUIT participants showed increases in five of the five career decision-making self-efficacy factors, including self-appraisal, occupational information-seeking, goal selection, planning, and problem-solving. No such changes, however, were observed across the same testing period for the control group. Regarding decision-making style, neither individuals in the INTUIT group nor individuals in the control group showed changes in the expected directions. Similarly, no significant results were seen on the TTFS measures from Time 1 to Time 2. Studies utilizing the TTFS resulted in a positive association with one’s tendency to foreclose and belief in career myths, but hypotheses indicating an association among TTFS scores and other work-related variables such as perceived barriers to employment, vocational identity, and need for occupational information were not supported (Ladany, Melincoff, Constantine, & Love, 1997). When INTUIT participants’ scores on the self-efficacy scales were compared at Time 2 with the control group, participants in the intervention group showed higher scores in self-appraisal and goal selection, but not in occupational information, planning or problem-solving. Hence, findings were mixed, but generally positive. The intervention appears to influence career decision-making through self-exploration. Most findings were in expected directions and interpretation was facilitated by the fact that scores for the INTUIT and control groups were relatively equal at pretesting.

The results of this evaluation suggest that traditional vocational outcome measures, such as the CDMSE-SF, may be useful in gauging changes in the career decision-making self-efficacy of female offenders preparing for career reentry. A prior group-based intervention for women used the CDMSE-SF to show that participants experienced an improvement in their career decision-making self-efficacy postintervention (Sullivan & Mahalik, 2000). Observed changes were small, but in expected directions. Few studies have documented specific career-related gains in offenders. Given that, for female offenders, employment has been associated with reduced recidivism rates (Greenfield & Snell, 1999), these gains may represent an important meter of progress in career development and successful reentry. Furthermore, the finding that INTUIT participants showed higher scores in self-appraisal and goal selection is likely due to the emphasis on formal career-related self-assessment.

In this Southeastern state, we could not allow offenders to conduct real searches using ONet Online* and other career-related, Internet-based resources. Hence, our findings may relate to difficulties in providing actual occupational information relevant to individual participants. The lack of gains related to planning and problem-solving may represent limitations in programming or, possibly, premature measurement of these constructs. Although our participants were under community corrections residential supervision in a Southeastern state, including work placement during the day, they were not yet facing actual concerns with regard to planning and problem-solving. Following participants upon their release from all correctional supervision, and measuring constructs during reentry seems a much better measurement strategy, and one in which positive results may more likely be seen.

Limitations

This program evaluation has clear limitations. First, and because the INTUIT intervention was developed with the chief goal of assisting female offenders, the project was not constructed as a formal clinical trial. This study lacks the sophistication of such an effort and, as just mentioned, longer-term follow-ups were not planned. Desistence from crime is the ultimate outcome measure in this area of research. However, we made no attempts to track participants past completion of the program. Hence, it is unclear whether the increases seen in our analyses translate into clinically meaningful change. Offenders’ life experiences following release would need to be observed to establish behavioral changes in career-related phenomena.

Additionally, the study can be faulted for the lack of strict random assignment of participants to condition. This problem is seen throughout the literature on offender reentry and in previous generations of the career intervention literature. Despite this, we argue that allocating participants to groups based on their admission to the facility––especially when data are collected over a 9-year span––should approximate randomness, and small but noteworthy outcomes were seen in traditional career development variables.

Future Directions

This article reports on the evaluation of a second-generation career intervention for nonviolent female felony offenders. Although grounded in SCCT constructs, the program was designed to maximize factors identified by Brown and Ryan-Krane (2000) as especially predictive of positive effects in career decision-making efforts. Of these features, the intervention seemed to help individuals to clarify their career goals and conduct a high-quality self-appraisal that likely was facilitated using individualized career-related assessments. As noted by Fouad (2007), self-efficacy and outcome expectations are important in that they predict interests which, in turn, predict choices and ultimately performance. Results of this study suggest that career-related decision self-efficacy can be modified in female offenders.

Because of the success of the INTUIT program, it was selected for inclusion in the State Council of Government’s Report of the Re-Entry Policy Council. As such, the program was said to exemplify several consensus-based recommendations for policymakers, and for practitioners who are interested in improving the likelihood that adults released from prison or jail will avoid crime. The innovative aspects of the program include (a) working with female offenders’ limited exposure to the legitimate labor force and using reflective tasks to address their sense of foreshortened futures; (b) helping them to explore the stigma of incarceration, and how to address the issue of a felony conviction during workforce reentry; and (c) formally identifying female offenders’ work-related values and interests using standard career-related assessment devices. The INTUIT intervention is complete, additional data are available for analysis, and the project remains a second-generation career intervention for nonviolent female felony offenders grounded in SCCT. The experience gained by developing and conducting this intervention, with advances seen in research and practice may lead to a postrelease program for offenders that can assist female offenders in their career-related reentry.

	Intervention	Control
Race
African American	36.4%	37.8%
Asian American/Pacific Islander	0.4%	0.0%
Caucasian	59.7%	58.4%
Hispanic	1.6%	0.8%
Middle Eastern	0.4%	0.0%
Other	1.6%	2.9%
Age in years at start of program	31.34 (7.91)	30.42 (8.07)
Marital status
Single, never married or partnered	22.2%	36.1%
Married or partnered	48.6%	36.6%
Separated	13.4%	13.0%
Divorced	14.2%	12.6%
Widowed	1.6%	1.7%
Average number of children	1.84 (1.62)	1.68 (1.40)
Highest educational achievement
Some high school	31.6%	31.9%
High school diploma/GED	36.4%	33.6%
Some college/2-year degree	16.6%	23.1%
College graduate	4.3%	6.7%
Other (indicated graduate degrees)	9.1%	3.8%

	Control Group	Intervention Group
Age	238	30.42	8.07	253	31.34	7.91	NS
Children	238	1.68	1.40	253	1.84	1.62	NS
TTFS	237	32.23	9.32	245	31.76	9.28	NS
MDMQ Buck Passing	238	14.87	2.68	252	14.36	2.75	NS
MDMQ Vigilance	238	8.76	2.31	251	8.94	2.31	NS
S-APP	238	19.48	3.72	253	18.76	4.94	NS
INFO	237	19.36	4.04	253	18.86	5.57	NS
GOAL	238	19.03	3.82	252	18.27	4.15	NS
PLANS	238	18.44	4.08	253	17.87	4.44	NS
P-SOLV	238	17.95	4.04	253	17.92	4.18	NS

	Time 1	Time 2
TTFS	229	31.61	9.19	30.55	8.55	NS
MDMQ Buck-Passing	252	14.36	2.75	14.02	4.11	NS
MDMQ Vigilance	251	8.94	2.31	8.09	2.83	<.001
S-APP	243	18.59	4.36	19.95	3.87	<.001
INFO	242	18.82	5.66	19.97	4.20	.001
GOAL	242	18.26	4.20	19.44	3.72	<.001
PLANS	243	17.86	4.54	19.31	3.96	<.001
P-SOLV	243	17.94	4.19	18.67	3.77	.008

	Control Group	Intervention Group
S-APP2	206	19.50	4.01	243	19.95	3.87	8.21	.004
S-APP1	206	19.48	3.72	243	18.76	4.94	131.03	< .001

	Control Group	Intervention Group
GOAL2	206	18.91	3.90	242	19.44	3.72	10.22	.001
GOAL1	206	19.03	3.82	242	18.27	4.15	155.06	< .001

Footnotes

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.

References

Andrews

L. M.

Bullock-Yowell

Dahlen

E. R.

Nicholson

B. C.

(2014). Can perfectionism affect career development? Exploring career thoughts and self-efficacy. Journal of Counseling & Development, 92, 270–279.

Bandura

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

Betz

N. E.

Klein

K. L.

Taylor

K. M.

(1996). Evaluation of a short form of the career decision-making self-efficacy scale. Journal of Career Assessment, 4, 47–57.

Blustein

D. L.

Ellis

M. V.

Devenis

L. E.

(1989). The development and validation of a two-dimensional model of the commitment to career choices process. Journal of Vocational Behavior, 35, 342–378.

Brown

Reedy

Fountain

Johnson

Dichiser

(2000). Battered women’s career decision-making self-efficacy: Further insights and contributing factors. Journal of Career Assessment, 8, 251–265.

Brown

D. S.

Ryan-Krane

N. E.

(2000). Four (or five) sessions and a cloud of dust: Old assumptions and new observations about career counseling. In Brown

S. D.

Lent

R. W.

(Eds.), Handbook of counseling psychology (3rd ed., pp. 740–766). New York, NY: Wiley.

Chartrand

J. M.

Rose

M. L.

(1996). Career interventions for at-risk populations: Incorporating social cognitive influences. Career Development Quarterly, 44, 341–354.

Di Fabio

Palazzeschi

(2012). Incremental variance of the core self-evaluation construct compared to fluid intelligence and personality traits in aspects of decision-making. Personality and Individual Differences, 53, 196–201.

Durose

M. R.

Cooper

A. D.

Snyder

H. N.

(2014). Recidivism of prisoners released in 30 states in 2005: Patterns from 2005 to 2015 (Special Report NCJ 244205). Bureau of Justice Statistics.

10.

Fitzgerald

E. L.

Chronister

K. M.

Forrest

Brown

(2012). OPTIONS for preparing inmates for community reentry an employment preparation intervention. The Counseling Psychologist, 41, 990–1010.

11.

Fouad

N. A.

(2007). Work and vocational psychology: Theory, research, and applications. Annual Review of Psychology, 58, 543–564.

12.

Greenfield

L. A.

Snell

T. L.

(1999). Women offenders. (Bureau of Justice Statistics Special Report NJC 175688). Washington, D.C.: National Criminal Justice Research Service.

13.

Hurry

Brazier

Parker

Wilson

(2006). Rapid evidence assessment of interventions that promote employment for offenders [Research Report Number 747]. London, England: Department for Education and Skills.

14.

Ladany

Melincoff

D. S.

Constantine

M. G.

Love

(1997). At-risk urban high school students’ commitment to career choices. Journal of Counseling & Development, 76, 45–52.

15.

Lawrence

Mears

Dubin

Travis

. (2002). Research report: The practice and promise of prison programming. Washington, DC: Urban Institute Justice Policy Center. Retrieved from www.urban.org

16.

Lent

R. W.

Brown

S. D.

Hackett

(1995). Toward a unifying social cognitive theory of career and academic interest, choice, and performance. Journal of Vocational Behavior, 45, 79–122.

17.

Lent

R. W.

Brown

S. D.

Hackett

(2000). Contextual supports and barriers to career choice: A social cognitive analysis. Journal of Counseling Psychology, 47, 36–49.

18.

Lent

R. W.

Brown

S. D.

Hackett

(2002). Career development from a social cognitive perspective. In Brown

Brooks

Associates (Eds.), Career choice and development (4th ed., pp. 255–311). San Francisco, CA: Jossey-Bass.

19.

Lopez

F. G.

(1997). Further validation of the Commitment to Career Choices Scale: Extensions to a high school sample. Journal of Career Assessment, 2, 398–407.

20.

Mann

Burnett

Radford

Ford

(1997). The Melbourne Decision-Making Questionnaire: An instrument for measuring patterns for coping with decisional conflict. Journal of Behavioral Decision Making, 10, 1–19.

21.

Mau

(1995). Decision-making style as a predictor of career decision-making status and treatment gains. Journal of Career Assessment, 3, 89–99.

22.

Petersilia

(2004). What works in prisoner re-entry? Reviewing and questioning the evidence. Federal Probation, 68, 4–8.

23.

Porter

A. C.

Raudenbush

S. W.

(1987). Analysis of covariance: Its model and use in psychological research. Journal of Counseling Psychology, 4, 383–392.

24.

Ross

Fabiano

(1985). Time to think: A cognitive model of delinquency prevention and offender rehabilitation. Johnson City, TN: Institute of Social Sciences and Arts.

25.

Shivy

V. A.

J. J.

Moon

A. E.

Mann

S. C.

Holland

J. G.

Eacho

(2007). Ex-offenders reentering the workforce. Journal of Counseling Psychology, 54, 466–473.

26.

Sullivan

K. R.

Mahalik

J. R.

(2000). Increasing career self-efficacy for women: Evaluating a group intervention. Journal of Counseling & Development, 78, 54–62.

27.

Varghese

F. P.

(2013). Vocational interventions with offenders interdisciplinary research, theory, and integration. The Counseling Psychologist, 41, 1011–1039.

28.

Varghese

F. P.

Fitzgerald

E. L.

Chronister

K. M.

Cummings

D. L.

Forrest

(2013). Vocational psychology with criminal justice populations: Why not? The Counseling Psychologist, 41, 1072–1082.

29.

Vigesaa

L. E.

Bergseth

K. J.

Jens

K. R.

(2016). Who participates in reentry programming? An examination of women offenders in a Midwestern state. Journal of Offender Rehabilitation, 55, 308–328.

30.

Visher

C. A.

Kachnowski

(2007). Finding work on the outside: Results from the “returning home” project in Chicago. In Bushway

Stoll

M. A.

Weiman

D. F.

(Eds.), Barriers to reentry? The labor market for release prisoners in post-industrial America (pp. 80–113). New York, NY: Russell Sage.

	Control Group			Intervention Group
	N	Mean	SD	N	Mean	SD	p
Age	238	30.42	8.07	253	31.34	7.91	NS
Children	238	1.68	1.40	253	1.84	1.62	NS
TTFS	237	32.23	9.32	245	31.76	9.28	NS
MDMQ Buck Passing	238	14.87	2.68	252	14.36	2.75	NS
MDMQ Vigilance	238	8.76	2.31	251	8.94	2.31	NS
S-APP	238	19.48	3.72	253	18.76	4.94	NS
INFO	237	19.36	4.04	253	18.86	5.57	NS
GOAL	238	19.03	3.82	252	18.27	4.15	NS
PLANS	238	18.44	4.08	253	17.87	4.44	NS
P-SOLV	238	17.95	4.04	253	17.92	4.18	NS

	Control Group			Intervention Group
	N	Mean	SD	N	Mean	SD	F	p
S-APP2	206	19.50	4.01	243	19.95	3.87	8.21	.004
S-APP1	206	19.48	3.72	243	18.76	4.94	131.03	< .001