Keywords

With the rise of US incarceration rates, there has been an inevitable rise in the number of incarcerated parents with minor children (Glaze & Maruschak, 2008). Bolstered by a growing literature on the impact of incarceration on children (see Chaps. 5 and 6 of this volume), as well as policies that link improved family relationships with reduced reoffending (e.g., H. R. 1593: Second Chance Act of 2007), institutions have responded by providing parenting training programs for incarcerated parents (Hughes & Harrison-Thompson, 2002). In a survey of key personnel from state correctional departments, Pollock (2003) reported that 38 of the reporting states had some form of parenting classes for incarcerated parents. Hughes and Harrison-Thompson (2002) gathered information directly from 315 participating state prisons and found that approximately half of the institutions offered parenting programs . While these data indicate that programming designed to provide parenting skills training is available in correctional settings, the numbers may overestimate the percentage of participants in those programs. Glaze and Maruschak’s (2008) survey of incarcerated parents revealed that only about 22–30% of mothers and 9–12% of fathers in state prison participated in parenting or child -rearing classes during their incarcerations.

To date, there have been a handful of literature reviews on this topic (e.g., Dallaire & Shlafer, 2017; Eddy & Burraston, 2017) and two meta-analytic reviews assessing the effectiveness of parent education programs for incarcerated populations, including an unpublished meta-analysis focusing on prison nursery programs by Shlonsky et al. (2016) and a meta-analysis by Armstrong, Eggins, Reid, Harnett, and Dawe (2017) of 16 empirical investigations. Both meta-analytic reviews indicate positive impacts of programming across different types of correctional settings (e.g., jails and prisons) despite the use of different program curricula. Armstrong et al. (2017) concluded that a small to moderate effect was found for increases in parent knowledge and improved quality of parent–child relations over comparison groups. Shlonsky et al. (2016) found that mothers who participated in prison nursery programs were less likely to recidivate than mothers who were separated from their newborns. Building from this work, we examine the impact of parenting programming for incarcerated fathers and mothers from a gendered perspective and consider the existing body of work in the context of research findings on parent programs and interventions that have been delivered and studied outside of correctional environments.

Unique Aspects of Parenting Programs for Incarcerated Parents

There is a rich and well-researched body of work on empirically supported approaches to parent training for behavioral problems in children (e.g., antisocial, noncompliant, aggressive, acting out behaviors) who come from a variety of populations (e.g., Dishion & Snyder, 2016; Sanders, Sanders, Markie-Dadds, Tully, & Bor, 2000; Thomas, Thomas, & Zimmer-Gembeck, 2007; Webster-Stratton, 2001). However, with only a few exceptions (e.g., Mindel & Hoefer, 2006; Palusci, Crum, Bliss, & Bavolek, 2008; Schiffmann, Eddy, Martinez, Leve, & Newton, 2008), this work has not been translated into programs for incarcerated fathers and mothers. More frequently, the existing interventions in jails and prisons have been drawn from “universal parenting programs ” (Valle et al., 2004) designed to provide broad psychoeducational support to parents within the general population. While there have been a few empirical investigations of some of these broader community -based programs, in general they have not received the intense level of empirical attention garnered for interventions designed for specific child behavioral problems.

The gap between interventions inside and outside of the carceral environment reflects the unique aspects of parental incarceration . Outside of jail or prison , parents typically seek parenting consultation because there is a specific problem with an identified child . Meta-analyses of parenting interventions (e.g., De Graaf, Speetjens, Smit, De Wolff, & Tavecchio, 2008) frequently focus on the reduction of child disruptive behaviors as a common marker of program success. The child ’s problem behavior creates an immediate motivation for participation in treatment , with the timing to start as soon as possible. However, for the incarcerated parent , motivation for intervention can reflect a myriad of goals for self-improvement and the timing for entry into classes is likely dictated by the availability of programs and the parent ’s eligibility for enrollment. Moreover, the typical skills covered in “outside” parenting interventions may not be immediately applicable. Learning how to handle children ’s tantrums and other misbehaviors may have a limited shelf life for the incarcerated mother who has no opportunity for practice .

Likewise, there are unique skills addressed in parenting programs for incarcerated parents that are largely irrelevant in parenting programs for the non-incarcerated. Many of the components often included in parenting interventions for incarcerated mothers and fathers—such as instruction in specific communication avenues (i.e., letter writing, phone calls, and personal visits during incarceration ), strategies for better collaboration with at-home caregivers, awareness of legal rights concerning children , and ways to deal with intense emotions regarding separation , loss, and incarceration —would not ordinarily have a place in interventions outside of the jail or prison .

Parenting Program Content and Parent Gender

Program Content

The content of interventions varies considerably, making it difficult to portray a “standard” parenting intervention . The quality and curricula of parent education programs also vary widely, and there are no “commonly accepted best practices for parenting education and skills training ” for incarcerated individuals (Eddy et al., 2008, p. 89). Eddy and colleagues found that in their survey of 41 state and federal facilities, although nearly all provided some programming related to communication skills and parenting techniques, far fewer emphasized anger and stress management, provided visitation opportunities, or offered education on child development.

Interventions also vary considerably in length, duration, as well as other features. For example, LaRosa and Rank’s (2001) Real Life Parenting Skills Program met for one-half hours once a week for five weeks. By contrast, Sandifer (2008) implemented the Rebonding and Rebuilding (Meyer & Moriarty, 1995) curriculum, which met for 3 h a day, twice a week for twelve weeks. Some interventions feature the inclusion of visit experiences (e.g., Snyder-Joy & Carlo, 1998) as integral to the training . Some interventions are aimed at parents nearing their release date from jail or prison (Bushfield, 2004; Maiorano & Futris, 2005), while others include parents who are facing long sentences (Loper & Tuerk, 2011). Comparisons based on the content of interventions thus become a comparison of myriad approaches. A more fruitful way to understand the big picture in parenting programs in prisons and jails is to look at commonality in targeted outcomes with a gendered approach that recognizes differences in parenting experiences for mothers and fathers.

Gender-responsive Programming

Parent education programs are sometimes specific to fathers (Antonio, Winegaurd, Young, & Zortman, 2009; Maiorano & Futris, 2005; Skarupski et al., 2003), to mothers (Harm, Thompson, & Chambers, 1998; Loper & Tuerk, 2011; Thompson & Harm, 2000), or both mothers and fathers (Eddy et al., 2008; Palusci et al., 2008). Some scholars have argued that the criminal justice system and correctional facilities should adopt and employ more gender-responsive policies that take into account differences between mothers and fathers. Covington and Bloom (2006) define gender responsive as “creating an environment through site selection, staff selection, program development, content, and material that reflects an understanding of the realities of the lives of women and girls and that addresses and responds to their strengths and challenges” (p. 9). In short, a gender-responsive approach assumes that gender makes a difference in parenting and takes into account a parent ’s identity as a mother or father when programming decisions are made.

Traditionally, correctional facilities were not designed with gendered needs in mind. Further, at least in principle, men and women involved in corrections are typically treated relatively equally with regard to gender. However, a gender-responsive approach puts gender at the forefront by acknowledging that gender is impactful within multiple settings and roles in day-to-day life. The National Institute of Corrections (NIC) began a gender-responsive project in 1999 at four pilot sites across the USA. The results of the pilot project suggest that when a gendered approach is applied at intake, in classification, and in programming decisions, the chances of successful reentry for women increase (Wright, Van Voorhis, Salisbury, & Bauman, 2012).

The gender of the incarcerated parent is central in the case of the incarceration of a pregnant or postpartum woman. Many jails and prisons are ill-equipped to deal with the medical and emotional needs of pregnant and postpartum women (see Ferstz & Clarke, 2012; Kelsey, Medel, Cumings, Dallaire, & Forestell, 2017). However, prison nursery programs, which are only available in a few correctional facilities in the USA , help to address the needs of these women and their babies and embody a gender-responsive approach (see Chap. 12, this volume). In our review, we include the findings from empirical assessments of parenting programs , including prison nursery programming, in separate tables. Table 13.1 presents the results of 16 empirical investigations of programs implemented with fathers. Table 13.2 presents the results of 22 empirical investigations of programs implemented with mothers. When studies included both father and mother participants, we included the study in both tables (e.g., Eddy, Martinez, & Burraston, 2013).

Table 13.1 Parenting programs for incarcerated fathers
Full size table
Table 13.2 Parenting programs for incarcerated mothers
Full size table

Unevaluated, Qualitative, and Quantitative Evaluations of Parenting Programs

In the USA , there are many parenting programs for incarcerated parents that generally fall into one of the three categories: (1) unevaluated interventions; (2) qualitative descriptions of ongoing programs with limited quantitative documentation; and (3) quantitative studies, typically pre-post designs, that statistically evaluate changes among parents after participation in the program. Whereas the current chapter focuses on quantitative investigations of parenting programs , this category represents only a small portion of what is actually implemented in prisons and jails today.

Unevaluated Parenting Programs

There are numerous unevaluated programs designed to improve parenting skills within US jails and prisons. Diverse programs may be implemented by a variety of outside organizations, religious groups, or community volunteers. Curricula may consist of informal lectures, discussions led by individuals from the community , religious discussion about parenting, or other supportive activities. The content is generally dependent upon the knowledge and experience of the volunteer trainer. Although these classes are usually welcome additions to programming offerings at correctional institutions, the diversity of offerings, trainers, and content precludes a systematic overview within the limits of this chapter.

A number of additional types of programs which address other relevant topics for incarcerated parents are also typically available. These include a diverse array of faith-based programs. At least one or more of these programs in most institutions is a 12-step (e.g., Alcoholic Anonymous) program. Other programs that are outside the parenting realm per se, but relevant, are cognitive skills training programs. Some of these have evidence to suggest that they have a positive impact in various areas of daily life, including problem-solving relevant to interpersonal relationships (e.g., Thinking for a Change; Bush, Glick, & Taymans, 2016).

Qualitative Evaluations of Parenting Programs

In the academic literature, many studies qualitatively describe parenting education programs for incarcerated populations (e.g., Bruns, King, & Stateler, 2003; Kazura, 2001; Meek, 2007; Robbers, 2005) or use qualitative means to collect information in addition to empirically generated results (e.g., Antonio et al., 2009; Bushfield, 2004; LaRosa & Rank, 2001; Skarupski et al., 2003; NFI, 2008). These studies typically use informal interviews to learn what participants find useful about the parenting programs offered, what is missing from the program, and ideas for improvements. For example, Meek (2007) collected course feedback through open-ended questions following a one-week intensive parenting class for 75 young fathers. When queried regarding the usefulness of various components of treatment , participants valued general childcare issues, such as the correct way to care for a child , and more specific issues related to physical care of children , such as learning how to change diapers. Areas that the participants felt were absent from the class varied widely depending on the individual. All participants rated the course in the “fairly” to “very useful” range.

Qualitative studies may aid in understanding how incarcerated mothers and fathers view themselves as parents and their attitudes toward parenting in general (Bushfield, 2004; Robbers, 2005). Robbers (2005) found that the most beneficial aspect of a 10-week program for fathers in prison was, reportedly, an increase in self-esteem and renewed desire to build relationships with children . Incarcerated parents also reported an increase in contact with their children as a result of the knowledge and confidence gained through the program. Bushfield (2004) reported that, after parenting training , fathers re-evaluated attitudes regarding the importance of involvement in their child ’s life. Generally, qualitative studies demonstrate that parenting education programs are met with approval from the participants. The wide variety of responses regarding optimal components of treatment suggests that while incarcerated parents find interventions useful, they have diverse needs that may be difficult to meet with a single program.

Quantitative Empirical Investigations

Quantitative studies that evaluate parenting programming in correctional settings are few in number. Tables 13.1 and 13.2 summarize our review of empirically based evaluations of parenting programs for fathers (Table 13.1) and mothers (Table 13.2). They include any evaluation that we could locate through a search of the PsycINFO and National Criminal Justice Reference Service databases, additional searches on the Internet, and the cross-checking of references of studies provided in each of the articles we located. In cases where insufficient information was available from these sources, we personally contacted key individuals to obtain unpublished reports or other information. In several instances, the results of an evaluation were articulated in state reports or contract summaries rather than academic journals. We included in our tables any study we could locate that included at least a pre-post design, a quantitative measure of an outcome of interest, and descriptive information regarding the specific parenting program approach used. The tables indicate whether a comparison or control group was utilized and whether there was random assignment to groups.

Typical outcomes of empirical investigations of parenting programs (listed in terms of frequency of use in currently reviewed studies) include: (1) knowledge and attitudes, defined as acquisition of information regarding child development and socially normative beliefs about appropriate child -rearing, discipline, and the role of a parent ; (2) mental well-being and parenting stress, defined as improvement in mood, self-image, and stress levels; and (3) behavioral changes, such as frequency of contact and communication with children , rate of recidivism, and reduction of negative or harmful behaviors (e.g., institutional misconduct, substance use ).

It should be noted that, in general, child outcomes are either not included in evaluations of correctional parenting programs , or they are obtained by reduced-sample auxiliary measures (Harrison, 1997; Landreth & Lobaugh, 1998). This is in marked contrast to most parenting interventions for non-incarcerated mothers and fathers for which child outcomes are a primary marker of program success. The major exception to this is in evaluations of prison nursery programs, which routinely examine impacts on newborn health , well-being and attachment (e.g., Byrne, Goshin, & Joestl, 2010), and in higher quality studies, such as the largest randomized controlled trial of a parenting intervention in a corrections system to date (Eddy et al., 2013).

Knowledge and Attitudes

The most widely used benchmark of a successful parenting education program is a significant change in attitudes or knowledge about parenting. Consistent with the meta-analysis conducted by Armstrong et al. (2017), all of the empirical studies listed in Tables 13.1 and 13.2 report pre-post improvement in at least one aspect of knowledge or attitudes. While the instrumentation varies widely, several studies used the Adult-Adolescent Parenting Inventory (AAPI-2; Bavolek & Keene, 2001). The measure was introduced in 1979 and has since been used across a variety of studies to assess change in parenting attitudes and knowledge (Family Development Resources, 2008). This self-report measure yields an overall score and five subscales that measure attitudes involving inappropriate parental expectations, empathy toward children , corporal punishment, parent –child role expectations, and child need for power and independence.

Palusci et al. (2008) used the AAPI-2 to measure the change in parenting attitudes and knowledge following the implementation of a parenting education program in a variety of settings within a community , including a local jail . The program, Helping Your Child to Succeed (HYCS), is a 10-week-long program in which the parents meet weekly with trained counselors and social workers. The curriculum was adapted from a universal parenting education program , the Family Nurturing Program (Bavolek, 1999), described as a “family -centered program proven to help parents and children learn to care for themselves and each other and to replace old, unwanted abusive interactions with newer, more nurturing ones (Family Development Resources, 2008).” The 10 sessions of HYCS are devoted to teaching 10 “democratic” child -rearing topics, such as positive attention and praise, setting appropriate expectations, and developing healthy communication patterns. Parents incarcerated at a county jail , in addition to other community members, participated in HYCS as a part of a 10-week substance abuse treatment program (Palusci et al., 2008). Of the parents who participated during a six-year span, 372 completed both pretest and posttest measures of the AAPI-2. Palusci et al. (2008) reported that mean scores increased significantly in a positive direction on four of the five constructs (e.g., expectations, empathy).

Other studies presented in Tables 13.1 and 13.2 use a similar design to Palusci et al. (2008) and used the AAPI to measure change (Bavolek, 1984; Bavolek & Keene, 2001). At first glance, this would seem to be a welcome sign and an opportunity to draw conclusions across programs using meta-analytic approaches. However, methods for the actual use of the measure vary substantially. For example, Robbers (2005) used only 7 of the 40 items, and Bushfield (2004) only reported scores for items with significant pre-post changes. Harrison (1997), like Palusci et al., drew from Bavolek’s Nurturing Program, but only reported on one AAPI score, and did not provide full descriptive information (e.g., scale standard deviations). Harm et al. (1998) likewise presented limited descriptive information regarding performance on all subscales. Thus, although there is welcome common measurement across several studies, and consensus that attitudes improved with intervention , the variations in measurement patterns preclude making statistically based generalizations regarding the impact of parenting interventions on attitudinal change.

Items on other non-standardized instruments utilized by some parenting interventions comprise broad statements to which the participant self-evaluates his or her own parenting skills. Example questions include “I know how to talk about my child ’s feelings and emotions,” “I can parent my children effectively from prison ,” and “I am confident about my parenting skills” (Antonio et al., 2009; Gonzalez Romero & Cerbana, 2007; Maiorano & Futris, 2005). Generally speaking, parents show increased confidence in their attitudes and knowledge when responding to these types of items. However, so-called meta-cognitive assessment of beliefs and knowledge is not the same as direct measurement , and it is not clear whether so-measured change represents true shifts in maladaptive attitudes or broader confidence that one’s attitudes—adaptive or not—are correct.

A number of empirical studies do not utilize a standardized measure of attitudes or knowledge and favor researcher-designed surveys. The National Fatherhood Initiative (NFI; www.fatherhood.org), an organization that develops and evaluates parenting resources and education programs for fathers, uses this method for the internal evaluation of their many programs including InsideOut Dad, a NFI program designed specifically for incarcerated fathers (Block et al. 2014; NFI, 2005, 2008, 2009). The program consists of 12 one-hour sessions that address topics such as ground rules, self-awareness, being a man, spirituality, handling emotions, relationships, fathering, child development, discipline, and fathering from the inside. In preliminary evaluations, the program was implemented in several correctional institutions in Maryland and Ohio (NFI, 2008, 2009). Participants answered 26 multiple-choice questions before and after the program, such as “Self worth is a term used to describe: (a) How a person feels about himself, (b) What a person thinks about himself, (c) Both the feelings and thoughts a person has about himself, and (d) Don’t know.” Mean scores for parenting attitudes and knowledge improved significantly across facilities. In a subsequent evaluation of the program, Block et al. (2014) found general increases in parenting knowledge. In summary, increases in parenting knowledge are a common marker reported across empirical investigations. The majority of studies listed in Tables 13.1 and 13.2 reported increases in knowledge and improvements in attitudes.

Well-being and Parenting Stress

Incarcerated men and women have high levels of mental health problems, well beyond that found in non-incarcerated samples (James & Glaze, 2006). Incarcerated women, in particular, have high levels of depression , borderline personality, and other emotional problems (Jordan, Schlenger, Fairbank, & Caddell, 1996; Warren et al., 2002). A large body of evidence links parenting stress, or high levels of concern regarding the roles and responsibilities surrounding parenting, with impaired parenting as well as with various mental health problems that may, in turn, impact parenting (Ortega, Beauchemin, & Kaniskan, 2008; Rodgers, 1998; Rogers-Farmer, 1999).

A focus on developing methods for controlling stress regarding parenting and improving general emotional reactivity about child -related issues is appropriate for many incarcerated parents. For many incarcerated mothers, separation from their children represents the most excruciating and enduring pains of incarceration (Arditti, Smock, & Parkman, 2005; Clarke et al. 2005; Hairston, 1991). Helping incarcerated parents to deal with this unique form of pain can give the opportunity to “make lemonade from lemons,” to the extent that such is possible, and develop skills while in prison or jail that can improve communication and understanding.

Loper and Tuerk (2011) developed a program for long-term incarcerated mothers designed to equip incarcerated mothers with coping strategies for dealing with the stress of separation and to improve communication patterns with children and caregivers. The program employs a manual for each mother that elaborates on each of the eight sessions. Where possible, the sessions are structured using materials presented via computer presentation software, videotaped vignettes of difficult situations, followed by small group discussions. Central to all of the sessions is a reference to a cognitive-behavioral strategy that inserts conscious evaluation of ongoing assumptions and emotional reactions. Using the acronym “MOM-OK,” mothers learn to “Mellow Out,” using brief breathing and relaxation strategies; use their “Mind,” to identify dysfunctional thoughts; counter negative thoughts with “Other” possibilities; and self-query “What is best for my child [Kid].” This strategy is infused throughout all eight of the sessions. For example, during the sessions that focus on dealing with child questions about why the mother is incarcerated, the incarcerated mother might be urged to replace the cognition “Her father put her up to this to shame me,” with “She is curious and wants to understand why things are this way.”

Loper and Tuerk (2011) evaluated the benefits of the program in terms of reducing parenting stress and other mental health difficulties, improving mother alliance with child caregivers, and changing frequency of mother-initiated contact through letters. Pre-post intervention comparisons documented improvements on the Parenting Stress Index (Abidin & Brunner, 1995), Parenting Alliance Scale (Abidin & Konold, 1999), the Brief Symptom Inventory (Derogatis, 1993), and the frequency of letter writing. When significant pre-post comparisons were reanalyzed controlling for the frequency of using the MOM-OK strategy, previously significant effects were no longer significant, suggesting that the strategy may mediate some positive effects. However, pre-post changes were generally not significant in comparison with a waitlist group.

Other interventions that have targeted emotional and personal stressors have focused on a parents’ personal sense of self-esteem or confidence in their ability to parent effectively. For example, Harm et al. (1998) found improvements in self-esteem among a group of incarcerated mothers with substance problems using the Nurturing Parent (Bavolek & Comstock, 1985) curriculum. Along similar lines, in a later study, the same authors (Thompson & Harm, 2000) found that improvements in self-esteem were more apparent among mothers who had some contact with children , emphasizing the importance of opportunities to practice skills in achieving the desired outcomes.

In general, the interventions that have examined mental health issues have found positive changes in parenting stress and sense of well-being. The question arises as to whether such positive changes can then be generalized to improved parent–child interactions. The challenge for these programs, as is the case for programs designed to improve knowledge and attitudes, is in affording practice opportunities (and direct measurement ) of the acquired skills that are intended to positively impact well-being and stress.

Behavioral Changes

Changes in behaviors regarding contact and communication with children and caregivers at home are included in several evaluations. Less frequently, evaluations target reduction in recidivism. Parent-child contact and communication patterns can change abruptly and dramatically when the parent is incarcerated. A majority of incarcerated mothers and a substantial portion of fathers reside with their children prior to incarceration (Glaze & Maruschak, 2008). However, during incarceration , parent–child contact is typically limited to letter writing, phone calls, and visits (see Chap. 10, this volume). Institutional policies and financial burdens further limit the number of phone calls and visitation opportunities, and the cooperation of the caretaker and child can alter the success of the contact .

Assessment of change in communication patterns typically relies upon parent self-report of frequency of phone calls, letter writing, and visitation (Antonio et al., 2009; Harm et al., 1998; Kennon, 2003; NFI, 2008, 2009). A few studies also seek to assess change in the quality in communication by querying about the presence of specific patterns, such as yelling at children and telling children they are loved (Czuba, Anderson, & Higgins, 2006; NFI, 2008, 2009). Antonio et al. (2009) evaluated behavioral changes following the 12-week parenting program , Long Distance Dads, using parent responses to twelve self-reports of specific parent behaviors, such as “‘…how often have you’ ‘…talked about events that are currently going on in your child ’s daily life’ or ‘…evaluated your child ’s physical needs’.” The participants were also asked how often they sent gifts, communicated via phone or letters, or requested visits . Pre- and post-program analysis showed that those who completed the program increased frequency of talking about events in their children ’s lives, sending gifts, phoning, and assessing their children ’s physical and emotional needs.

In general, evaluations of programs report mixed results concerning changes in contact frequency, with some investigations showing improvement (Antonio et al., 2009; Loper & Tuerk, 2011; Robbers, 2005; Skarupski et al., 2003) and others not detecting change (Gonzalez et al., 2007; Kennon, 2003). Null results may reflect insufficient focus on this outcome, variations in institutional constraints regarding contact , lack of family resources, and other unmeasured variables.

Mindel and Hoefer (2006) evaluated change in parental behaviors following a family strengthening program offered through a substance abuse treatment facility both for parents who were nearing release or who recently released and for their children . This was one of the few studies we review in this chapter that implemented a curriculum adapted from a universal parenting program that met the criteria as an evidence-based program by the former National Registry of Effective Programs and Practices. The 14-week program included separate 60-min meetings for children and parents, followed by a communal meeting to provide opportunity for the parents to practice newly learned skills. Mindel and Hoefer’s study is exceptional in its inclusion of measures completed by participating children . Incarcerated parents as well as their children reported improvements in family bonding and parental involvement, as well as an increase in the opportunities and rewards that come with prosocial behavior.

The rationale for educational opportunities in correctional settings rests on the assumption that such intervention reduces the likelihood of dysfunctional behaviors that lead the parent to reoffend after release. Parenting education may reduce conflict and stress with family members that result in a more successful adjustment during and after prison , which in turn reduces offending . Very little research has evaluated the impact of parenting programming on recidivism, and what exists yields little support for the assumption. Maiorano and Futris (2005) found that while recidivism rates declined slightly among fathers who completed a parenting program, they were comparable to the recidivism rates of the general prison population. Similarly, Gat (2000) found no significant differences in recidivism between a group of participating mothers and mothers who did not participate in the parenting program . While this objective makes sense in a correctional context, it may be overly optimistic to expect that relatively brief parenting interventions alone will be sufficient to reduce reoffending. Rather, the success of parenting education in reducing recidivism is likely better estimated in terms of a tandem operation providing additional forms of support for incarcerated parents and their families during and after incarceration (see Chap. 15, this volume).

Recurrent Limitations in Assessments of Parenting Programs for Incarcerated Parents

While relatively few in number, most of the publicly available reports regarding empirical evaluations of parenting programs in jails and prisons have been positive. However, several limitations appear across these studies. Much existing data-based research relies on pre-post designs rather than randomized designs with control groups. Of the 38 studies presented in Tables 13.1 and 13.2, over 65% (n = 25) involve the use of control or comparison groups; in most cases, the comparison groups are very limited in terms of size and composition. While the generally positive observed pre-post changes are encouraging, it is important to know whether these changes are independent of factors such as preexisting group differences, regression to the mean, and/or unmeasured environmental effects at the prison . In terms of this last issue, seasonal changes, proximity to holidays, large transfers of inmate populations, and changes in administration are but a few of the overarching agents of change in the attitudes and behaviors of incarcerated populations. Documentation that observed positive changes occur irrespective of systemic effects is particularly important for this environment.

Randomized control trials (RCTs) are currently considered the “gold standard” for evaluating the effectiveness of a psychosocial intervention (Donaldson, 1998). Due to the numerous difficulties in using this type of design within correction settings, it is unsurprising that relatively few RCTs exist for parenting programs for incarcerated mothers and fathers. The ones that do exist warrant mention; there are seven RCTs listed in Tables 13.1 and 13.2 (three for fathers, three for mothers, and one for mothers and fathers). Work by Eddy and colleagues (Burraston & Eddy, 2017; Eddy et al., 2008, 2013, 2018; see Chap. 15, this volume) on the Parenting Inside Out (PIO) parent management training program has demonstrated, and using an RCT design, that program participants (mothers and fathers) showed reductions in stress and depressed mood relative to participants in the control group, and program participants had lower recidivism rates (as measured by police arrests) than control participants. This RCT is strong in terms of design, sample size, measurement , and program efficacy. The number of participants in the trial (N = 359) was higher than in all of the other RCTs presented in Tables 13.1 and 13.2 combined. Parents with children of a certain age were targeted. Attrition in the study was low. Replications of studies with similar design characteristics on parenting programs for incarcerated mothers and fathers are very much needed.

Unfortunately, most investigations do not employ random assignment, and regardless, substantial dropout rates in the various studies that are available have been common. For example, attrition of approximately 50% of the initial sample was observed by Czuba et al. (2006), Loper and Tuerk (2011), Sandifer (2008), and Skarupski et al. (2003). The presumed initial equality of groups that is the objective of random assignment can be lost when significant portions of either group drop out. Moreover, institutional conditions may limit who is allowed to be part of a control group. For example, the control group in Antonio et al.’s (2009) evaluation of Long Distance Dads was comprised substantially of men who were ineligible for the training program due to problematic offenses, legal barriers to child contact , and lack of desire for program participation. These problems create substantial difficulties in understanding who is being evaluated and therefore to whom the intervention appropriately applies.

Many of the reported evaluations have very small sample sizes, sometimes due to high dropout rates described above (Browne, 1989; Bushfield, 2004; Gat, 2000; Harrison, 1997; Landreth & Lobaugh, 1998; LaRosa & Rank, 2001). The small sample sizes are particularly problematic for small waitlist comparison groups, as null results may be due to low power rather than lack of intervention effects. For example, after experiencing a considerable attrition rate, Sandifer’s (2008) evaluation of the Rebonding and Rebuilding curriculum was hampered by a small control group (n = 26). In several areas, the treatment group showed positive pre-post intervention gains, while the control group generally did not change on measured variables. While these results are encouraging, the observed absence of significant change in this waitlist group may reflect lowered statistical power to detect such changes. Further, waitlist attrition is a particular problem in corrections settings as incarcerated individuals may be transferred, experience incompatible schedule changes, commit infractions that restrict educational opportunity, or simply lose interest. It is not surprising that many interventions opt for simple pre-post designs rather than dealing with the likely difficulties of finding durable control groups.

By and large, most studies in the field have relied exclusively on self-report measures. In some cases, the measures reflect a self-evaluation of a quality rather than a more direct measure of the quality itself. For example, Robbers’ (2005) assessment of improved legal knowledge included the item “I know who to call to have my support payments adjusted if my employment status changes (p. 17),” rather than a direct query regarding who the inmate would call. The problems of using self-report are particularly risky with researcher-developed surveys that have not been subjected to psychometric scrutiny.

Why Are High-Quality Assessments of Jail- and Prison-Based Parenting Programs so Hard to Do?

The spotlight on the common limitations that so frequently plague parenting education programs in correctional settings leads to the question, “Why are there so few interventions that satisfy conditions that would be seen as fairly basic to evaluation of psychosocial intervention ?” The resounding answer is: “It’s a prison ,” or “It’s a jail .” There are numerous unique logistical, political, and practical considerations in conducting treatment or evaluation in a correctional environment that are not apparent in other settings. Some of the most basic needs for consistent programming—dependable location for training , reliable equipment, availability of materials—can be road blocked in a prison or jail . Delays in twice-daily person counts routinely cut into scheduled time. Lockdowns that interfere with holding a class are not uncommon. Unexpected transfers of class participants can result in dramatic changes in class size. Although the use of computer presentations is normative in most educational settings, jails and prisons often have restrictions on the use of computer equipment that preclude such innovation. Simple features such as turning on electric lights, rearranging furniture, and permitting small group discussion can be curtailed depending upon institutional security policies.

While concern for the well-being of the children of incarcerated parents is typically one purpose of education initiatives, few studies incorporate child outcome measures (Harrison, 1997; Landreth & Lobaugh, 1998). However, access to children is often very difficult in correctional settings. Families of the incarcerated irregularly visit their incarcerated loved ones, and the hospitality of conditions may vary widely (Kazura, 2001; Laughlin, Arrigo, Blevins, & Coston, 2008; Chap. 10, this volume). Many institutions have policies that prevent physical contact and limit communication during visitation . For example, incarcerated parents may be required to sit in a separate chair and refrain from touching. Long-distance travel to institutions can be burdensome on financially strapped home caregivers. The various personal activities of the inmate’s child —often scheduled on weekends—can interfere with weekend-only visitation hours. These scenarios make it difficult to adequately assess whether inmate parents are using targeted communication skills. Further complicating this is the fact that participants’ children vary widely in age. Parenting education and information that would be relevant for a mother or father of an infant may be less relevant for the parent of a teenager. Rarely do parenting programs screen or target participants whose children are between a specified age range.

Policy concerns can also influence how parenting interventions are devised and assessed. To implement a program in a correctional institution, it is sometimes necessary to demonstrate that the program has higher-order social benefits, beyond those for the individual family . For example, Antonio et al.’s (2009) evaluation of Long Distance Dads included goals “to become emotionally, morally, spiritually, psychologically, and financially responsible parents…” (p. 9). Along similar lines, Robbers’ (2005) intervention included objectives to: “Promote emotional, moral, spiritual and financial responsibility for children ” (p. 7). Many of Robbers and Antonio et al.’s other goals included objectives for skill development that are more typical in parenting interventions on the outside. However, the inclusion of goals for improved moral behavior would rarely, if ever, occur in interventions with the non-incarcerated.

This type of conceptualization of intervention can be important in gaining political support for the intervention . A survey of 200 citizens living in Florida, Applegate (2001) found that citizens were often skeptical about the provision of many of the possible services and amenities for incarcerated individuals. However, nearly all of the respondents indicated that they would be willing to support such programming if there was a clear linkage between the service and reduced offending . The provision of services in correctional settings can require selling the public and policymakers on the redemptive value of an intervention in ways that would not be otherwise needed on the outside. However, the focus on these objectives may obscure goals for acquisition and measurement of more parenting-specific skills. Furthermore, it would be challenging to assess the impact of a program on less tangible outcomes like moral responsibility.

Future Directions for Improving Parenting Intervention Scholarship

Despite the common limitations as well as the ubiquitous difficulties of conducting research in prison and jail settings, there has been a welcome increase over the past decade in the number of published evaluations of parenting interventions , as well as increased understanding of the value of such approaches. As given in Tables 13.1 and 13.2, we observed 38 publicly available quantitative evaluations of parenting programs between 2010 and 2017, in contrast to only 7 such studies between 1989 and 1999, and 23 between 2000 and 2009. In 2007, the US Congress passed the Second Chance Act, which provides localities with funding for initiatives to reduce prison reentry and specifically prioritizes interventions aimed at improving family relationships of prisoners. In Applegate’s (2001) public opinion survey regarding correctional services, over 90% of the respondents indicated support for psychological counseling as well as opportunities for family visit experiences.

Another optimistic sign is the presence of new initiatives that, while still in development, offer promise. For example, Eddy, Martinez, Schiffman and associates’ (2008) development and evaluation of a broader program that includes Parenting Inside Out but expands to address other key factors related to parents and families both during and following incarceration seems quite promising (Chap. 15, this volume). The difficulties and limitations of doing intervention in correctional settings will not change. However, knowledge can still grow by the adoption of several simple initiatives that would improve this important growing body of research .

Consistency and High Standards for Measurement

To demonstrate the value of parenting programs interventions, there needs to be a stronger and more unified effort regarding measurement of effects. More consistent use of established standardized measures, full reporting of descriptive information, and assessment of scale reliabilities within studies would enable opportunities for improved evaluation at little additional cost or effort. Furthermore, there is a greater push within all scientific disciplines, including psychology, for an “open science” defined as “the publication of scientific concepts together with the protocols and data upon which those concepts are based readily accessible to all levels of an inquiring society” (Hesse, 2018, p. 126). This push for open science emanates in part from the failure to replicate key findings, something which has plagued the field at large. Researchers who engage in the study of the children of incarcerated parents and their families could help achieve the goal of consistently high standards for measurement if more of us engaged in an open science way of doing business. Furthermore, another key problem with research on prison -based educational programs is that many unpublished studies are conducted by states and departments of corrections. Open science platforms could be made accessible to individuals conducting research in academic as well as non-academic settings, bringing a broader set of voices to the table to help move the work forward.

Exploration of Key Components of Change

There is also a need for better identification of the components of treatment that are presumed to mediate effectiveness. Most of the existing researches place emphasis on demonstrating that the approach works in improving some skill or belief. But if this is effective, why is the approach working, and how can mediating mechanisms be evaluated? While some qualitative studies explore this question by querying participants on useful program aspects, quantitative investigation of such mechanisms is largely lacking. Loper and Tuerk’s (2011) finding that high levels of the usage of the “MOM-OK” cognitive-behavioral strategy were associated with the observed improvements supports the inclusion of this feature as an important component. Attention to understanding the specific program mechanisms for change is needed for the continual revision and improvement of programming.

Inclusion of Child Measures

There is an obvious need for better documentation of the impact of programming on children . The types of information collected may well vary and should be consistent with program goals. For example, if an objective of a program includes teaching parents to be more sensitive to children ’s feelings and emotions regarding painful separation from parents, it would be useful to gain information about changes in the child ’s comfort level with the separation . This might be obtained by caregiver ratings, projective examination of child drawings, or self-report in interviews or simple measures. If the objective is to teach better child management techniques, behavior rating scales completed by caregivers or teachers could be useful. Although the quality and type of information collected will likely vary, it makes sense to gather this information for ongoing improvement of the intervention . Furthermore, there needs to be a greater understanding of how a parent -based educational program could impact a child in this type of setting, when a parent is typically limited in their interactions and contact with the child . At a minimum, a start toward the inclusion of child measures would be for researchers to report demographic information about the children of participants in corrections-based parenting programs (e.g., age, frequency of contact , living situation before incarceration ).

Opportunities for Practice

The inclusion of structured visitation programs that allow for practice of newly learned skills affords the opportunity for better acquisition and measurement of targeted skills. While some interventions include children through planned regular visitation programs (Landreth & Lobaugh, 1998), logistic and security features of many prisons bar this as a common practice . However, in these cases, more attention to direct measurement of skills can still be accomplished with a bit of creativity. Examination of letters sent before and after instruction on optimal written communication, for example, can afford objective information for assessment and instruction. Daily checklists in which parents monitor their use of covered strategies can provide routine information regarding treatment compliance. In-class exercises that call for actual practice of skills (e.g., role plays, observation, and critiques of video vignettes) afford opportunities for “virtual” practice as well as for simple measurement of skill acquisition.

Improvement in Documentation and Description of Treatment

Efforts to replicate and build upon the existing literature will require more detailed and comprehensive documentation of treatment content, curriculum, and implementation. Currently, there is wide variation in the level of the description provided for interventions, and limited information regarding the training or professional skills of the program facilitators. Manuals or documented guidelines for how to conduct sessions are rarely provided and may not even exist for some programs. There are, however, welcome exceptions to this pattern (Antonio et al., 2009; Czuba et al., 2006; Loper & Tuerk, 2011). Some programs use portions of the existing outside programs that provide documentation of training procedures. For example, Harrison (1997) used a combination of Bavolek & Comstock’s (1985) The Nurturing Program as well as components from Dinkmeyer and McKay’s (1989) Systematic Training for Effective Parenting. Specific descriptions of intervention and training content allow for the replication of reported successful interventions and guidelines on important intervention features. Further, and quite importantly, a need remains not only for the ongoing monitoring of implementation fidelity, but the documentation of such in published studies.

Delineation and Description of Contextual Features

The varying content of interventions likely reflects the various contexts in which intervention is implemented. Better attention to and description of these contexts will improve understanding and cross-fertilization of efforts. While there are many situations that characterize a particular setting or program, there are two major contextual dimensions that can substantially affect the content of programming. The first concerns whether the program is implemented for mothers or fathers. While a few interventions have been used with both men and women, many are specifically designed for mothers or fathers, or at least they are noted to be such. This is not surprising: Prisons are gender-specific, and jails have gender-specific sections. The needs and stresses of incarcerated mothers can differ considerably from men, due to differences in pre-incarceration primary caretaker status, length of sentence, connection with caretakers, presence of mental health problems, and many other gendered differences (Loper & Tuerk, 2011). Detailed descriptions of exactly how a program addresses specific aspects of being a mother or a father are needed.

Along similar lines, program content may vary depending upon whether reunification is expected within the short term or long term. Parents who will soon be resuming contact with children , as is the case in many jail programs, may benefit from more instruction in behavioral management and awareness of transitional issues that can arise with unification. Parents serving longer sentences may need more instruction regarding ways of utilizing the existing communication avenues, growing personally, and collaborating with caregivers. Unlike most empirically supported family interventions that specify a particular child issue (e.g., ADHD, conduct disorder, autism), interventions with the incarcerated may be better summarized in terms of the key contextual features that permeate the incarcerated parent –child relationship.

Implications for Practice and Policy

Practice Recommendations

This review of 38 parenting programs for incarcerated mothers and fathers illuminates several opportunities for practice and policy . First, for practitioners—individuals facilitating and implementing programs in correctional facilities—it is critically important to meaningfully assess and document both the program curriculum being used and to evaluate outcomes on an ongoing basis. Program content must align with learning objectives. Practitioners can partner with university researchers to assist with evaluation efforts to help ensure that program goals are linked to measurable outcomes and to see whether such outcomes are actually achieved. When decisions need to be made by administrators about services in light of budget constraints, having assessments of program impact may help save helpful programs.

This field, on the whole, would benefit from the use of an agreed upon set of evidence-based and evidence-informed “best practices” for program content, delivery, and evaluation relative to parenting programs in general. This would help guide the development of new programs. Practitioners are advised to select the best program for their population with consideration of the gender of the incarcerated parent , the likelihood the parent will reunify with the child (ren), and the setting of the program (e.g., jail or prison ). When possible, child age should also be a consideration. An effective program in one context should not be considered the answer to all contexts. For example, although a program with fathers incarcerated at a prison facility may show positive results, it may not be a good program for mothers incarcerated in a local jail .

Policy Recommendations

A major policy issue that impacts the type of programming and content of program relates to the gender of the incarcerated parent . Traditionally, prisons and jails, and to a certain extent educational programming within the correctional context, were all designed for men. However, increasingly, women are represented in correctional populations. Unfortunately, many correctional facilities do not yet have gender-responsive policies. A gender-responsive approach takes into account the incarcerated individuals’ gender during all aspects of criminal justice involvement, from intake, to programming decisions, to reentry and reunification support. A gender-responsive policy toward educational programming would acknowledge the fact that mothers and fathers often have different roles in their family and may have quite different histories of communication and interactional styles with their children , in part due to these different roles.

A second policy recommendation concerns visitation and opportunities for contact with the incarcerated mother or father during their incarceration and perhaps while participating in educational programming. For a parenting program to be effective, a parent needs to have an opportunity to interact with their child and practice newly acquired skills and behavioral responses. Policies that facilitate parent–child contact during incarceration would include child -friendly visiting rooms and policies and perhaps provide extended contact visits for children with their parents. Additional opportunities for enhanced connection concern the availability and affordability of phone calls to and from correctional facilities. Gender-responsive policies and child -friendly visitation policies would be major changes for many correctional facilities. However, these changes, along with institutional support for parent educational programming, could have benefits for the parent and their child and family in the long term.

Conclusion

Many prisons, jails, community agencies, and citizens have responded to the need to provide incarcerated parents with parenting programs . Although a relatively small portion of incarcerated parents in the USA are currently enrolled in these efforts, there appears to be a growing awareness of the importance of this type of intervention with this high-risk segment of the population. There is a need for more and better evaluation of parenting programs for incarcerated mothers and fathers, as well as a tolerance for the unique challenges of doing research and evaluation in correctional settings. Recent legislative attention to the needs of incarcerated individuals in the USA is a welcome sign. There appears to be growing support for aiding families affected by incarceration . Although there are difficulties in doing this work, there is plenty of room for the community of clinicians, community organizers, correctional professionals, and scholars to create and refine programs on the inside that make a difference on the outside.