Poverty exhibits direct and indirect effects on families’ social, emotional, and physical health and well-being, disproportionately affecting the capacity to address basic needs and access healthcare and education as well as overall morbidity and mortality (Evans & Kim, 2013; Johnson, Riis, & Noble, 2016). The burden of income inequality, coupled with the chronic stress associated with resource-poor environments (Cheng et al., 2016), leads to increased challenges related to parenting for families living in poverty as compared to their higher income counterparts (Brooks-Gunn & Duncan, 1997; Jocson & McLoyd, 2015; Pascoe et al., 2016). Just as low-income populations are denied access to buffering supports, parents living in poverty may lack access to those resources which specifically support positive parenting, such as the ability to spend time with their children, procurement of childcare and transportation, and access to parenting information and social support (Johnson, Riis, & Noble, 2016; Kaiser & Delaney, 1996). Parents living in poverty are more likely to face higher rates of psychological distress, substance abuse, and intimate partner violence (Murali & Oyebode, 2004). The literature has indicated that parental stress resulting from economic hardships also can decrease the quality of parenting (e.g., Conger, Conger, & Martin, 2010; Ellingsen, Baker, Blacher, & Crnic, 2014). For instance, chaotic neighborhood conditions are associated with harsh and inconsistent discipline and lower warmth (Conger et al., 2010; Jocson & McLoyd, 2015), and overall poverty with lower rates of cognitive stimulation and parental supervision (Murali & Oyebode, 2004). These outcomes, in turn, place children at risk for socioemotional and cognitive difficulties throughout development and into adulthood (Duncan et al., 2010; Luby et al., 2013; Pascoe et al., 2016), as outlined previously.

Despite the harmful effects of poverty on families, research has demonstrated that promoting positive parenting can reinforce the resilience of low-income parents. The literature has highlighted the importance of promoting learning and caregiver–child interactions for families overall and those living in poverty (Kaminski, Valle, Filene, & Boyle, 2008; Thompson, 2016), and has indicated that a focus on early caregiving is especially significant (Luby et al., 2013). Although the literature has not indicated a single parenting behavior that is most important, a number of strategies have been associated with positive child outcomes. For example, a meta-analysis by Kaminski et al. (2008) has supported the efficacy of parenting programs in preventing and improving child behavior problems. Specific findings have indicated that the inclusion of positive parent–child interactions, emotional communication skills, and parent use of timeout and consistency in the curriculum were strongly associated with larger program effects. A more recent review of the literature by Mesman, van IJzendoorn, and Bakermans-Kranenburg (2012) has indicated significantly lower levels of parental sensitivity among ethnic minority parents stemming from increased social and economic burdens and the importance of both reducing family stress and strengthening parental sensitivity. Seminal studies by Baumrind (1966, 1967) have noted the importance of parental control, nurturance, and maturity demand, all concepts which she illustrated were associated with improved child development and continue to serve as a framework for parenting research today. The literature also has emphasized the importance of developmental stimulation during early childhood, which has been associated with later child outcomes such as improved language development and school readiness (Baker & Iruka, 2013; Walker, Chang, Powell, & Grantham-McGregor, 2005). As these studies have revealed, parenting behaviors such as reciprocity and sensitivity, parental control, developmental stimulation, commitment to parenting, nurturance, and maturity demand are associated with positive parenting and improved child outcomes.

The Legacy philosophy prioritizes three tenets: Child outcomes depend upon the quality of the relationship between mother and child, successful parenting comes in many different forms, and—regardless of their environment—mothers hold the power to positively influence their child’s development. The five overall goals of Legacy are to (a) promote maternal responsibility, maternal investment, and maternal devotion of time and energy; (b) promote responsive, sensitive mother–child relationships; (c) support mothers as guides to their children’s behavioral and emotional regulation; (d) promote each mother’s facilitation of their children’s verbal and cognitive development; and (e) promote mothers’ sense of community. These goals apply to parents of all socioeconomic backgrounds; however, mothers affected by poverty may lack the opportunity to safely discuss and explore parenting issues in a supportive environment, observe positive role models who demonstrate sensitive and responsive parenting behavior, or experience positive support from others. The group sessions afford mothers these opportunities, with the support and guidance of the group leader. The Legacy logic model (Figure 1; Perou et al., 2012) describes these goals, the conceptual pathways through which they affect the positive parenting practices of interest in this study, and the ultimate intended child development outcomes.

The Miami and Los Angeles sites each developed and implemented their own curricula based on the Legacy model and the needs of their respective communities. While they share the same foundation and many similarities, the resulting interventions also showed some differences as a result of each site’s community, demographic, and cultural characteristics. Both sites tested and developmentally sequenced curricula adhere to the five program goals and core model components; were based on the same evidence from successful existing intervention studies; and repeatedly covered themes which the literature has identified as key constructs in early childhood development, such as discipline, attachment, developmental milestones, parenting stress management, establishment of goals and dreams for their children, and early literacy. These curriculum themes evolve as children develop; for instance, sessions focused on discipline move from mothers discussing how daily routines and rituals guide their children’s behaviors to mothers encouraging their children’s cooperation through praise and modeling, and eventually to mothers supporting positive behaviors and addressing inappropriate behaviors using consistent feedback such as praise and timeout. Sites differed in how they sequenced and used discussion and activities to introduce and reinforce concepts presented in the curricula. Sites also differed in the length and timing of the intervention. The Los Angeles curriculum is offered prenatally through child age of 3 years, and the Miami curriculum is offered from birth through child age of 5 years.

For both sites, Legacy groups cluster tightly around the age range of children to follow the developmentally sequenced curricula and create a shared experience among mothers. Trained facilitators deliver regular group meetings of mothers; components include mother-only time and mother–child time with a targeted “Legacy child.” Mother-only time focuses on discussion of topics to promote sensitive caregiving and understanding of developmental stages, and building social connections between the mothers. During mother–child time, group leaders support and model positive parent–child interactions and allow mothers to practice skills learned during mother-only content. In Miami, each group session included a mother-only and a mother–child component; in Los Angeles, group sessions alternated between mother-only and mother–child sessions. Group leaders present information in a nondidactic manner and respond to and reflect on each individual mother’s growth and needs. Both curricula also include community-building events and one-on-one time between mothers and the group leader to reinforce curriculum content. Perou et al. (2012) provided additional information on the Legacy model and curriculum delivery.

In addition to site differences in curricula, there also were some differences in the sample they served. Although both sites used the same basic eligibility criteria (e.g., poverty, English-speaking maternal primary caregivers; see Perou et al., 2012), demographic comparisons conducted during the original RCTs had indicated that mothers in Los Angeles were significantly older, more educated, and more likely to speak a language other than English in the home, as compared to mothers in Miami. In addition, mothers in Los Angeles were more likely to be married, Hispanic, and employed (Kaminski et al., 2013). Therefore, site differences in outcomes cannot be separated from both the differences in implementation/curriculum and sample characteristics. This study will examine mothers’ responses separately by site and provide additional context regarding some of these differences. This study will support interpretation of participant experiences through examination of mothers’ own words.

The current study examines how mothers participating in Legacy describe parenting following involvement in the program, particularly across implementation sites in Miami and Los Angeles. Specifically, we wanted to explore how mothers in Miami and Los Angeles describe knowledge and use of positive parenting practices, and compare the nature and content of discussions in both sites. Our intent at the outset was not to combine the distinct populations in Miami and Los Angeles but to better understand how different samples could benefit from the program. We focused our study around positive parenting practices frequently cited as effective within the literature and targeted by Legacy. Using these practices as a framework, we analyzed data from Legacy focus-group discussions to examine how mothers describe their commitment to parenting, nurturance of and sensitivity and responsivity to their child, parental control, maturity demands, and practices that encourage developmental stimulation (for a description of these practices as defined for this study, see Table 1).

The current study is a secondary analysis of focus-group data collected alongside the original RCTs. For the original study, the CDC contracted with RTI International to collect and broadly analyze focus-group discussion data (RTI Project No. 0206030.011); this initial exploration is summarized next and is described in more detail elsewhere (Fraser, 2009). The institutional review boards (IRBs) of the CDC; RTI; the University of California, Los Angeles; and the University of Miami reviewed and approved the focus-group protocol and discussion guides.

From 2005 to 2008, RTI researchers conducted a total of 21 focus groups with Legacy participants (n = 166), with 13 groups in Miami (n = 110) and 8 groups in Los Angeles (n = 56). Researchers created discussion groups based on levels of attendance and engagement in the program,¹ as well as recentness of program completion (for the frequency of participants by wave and site, see Table 2). RTI utilized purposive sampling, which allows researchers to recruit information-rich individuals (Hennink, Hutter, & Bailey, 2011). Participants provided informed consent, and each received transportation and $50 in recognition for their time and contribution. A trained moderator utilized a semistructured, pilot-tested guide to lead hour-long discussions. The guide focused on mothers’ reasons for participating in Legacy, aspects of the intervention that contributed to mothers’ sense of community, mothers’ motivation to attend group sessions (including facilitators of and barriers to attendance), and mothers’ perceptions of learned knowledge about parenting and child development. The guide also included follow-up probes to more fully capture participants’ perspectives. A second staff member compiled detailed notes and audio-recorded the discussions. Staff de-identified and transcribed the discussions verbatim. RTI International produced a final report with results of broad analyses as well as the participant characteristics described next (Fraser, 2009).

The current study focuses on positive parenting practices (commitment to parenting, nurturance/sensitivity/responsivity, parental control, maturity demand, and developmental stimulation) as they connect to the Legacy goals regarding the relationship between the mother and child (see Figure 1) (for a summary of the themes related to each positive parenting practice as described in this section, see Table 3).

Overall, Legacy mothers participating in focus-group discussions demonstrated understanding and use of positive parenting practices targeted by the Legacy goals. Mothers’ discussions clearly indicated their commitment to parenting—mothers described focusing on their child and the importance of staying engaged in their roles as parents. Mothers at both sites reported prioritizing their children and trying to become better parents. In addition, mothers’ statements also reflected nurturing cognitions and behaviors. Mothers reported wanting to be involved in their child’s life, learning to have more patience, and having concern for their child’s well-being. Mothers’ descriptions also reflected high levels of sensitivity and responsivity; they described talking and communicating more with their children versus before Legacy, improving in their understanding of their children, and realizing how their own responses affect their children. Mothers particularly highlighted the importance of physically getting down to their child’s level while communicating and reported seeing consequential improvements and growth in their children. Furthermore, mothers described improvements in their levels of parental control. Mothers explained that they felt more control of themselves and their children, described behavior change related to trying not to spank, and demonstrated increased knowledge of other disciplinary techniques. In addition, mothers expressed developmentally appropriate levels of maturity demand. They reported clear expectations for their children, but also flexibility regarding these expectations, granting their child a level of control tailored to their children’s ages, needs, and limits. Finally, mothers’ discourse revealed knowledge and use of practices to increase developmental stimulation (reading mostly, but other activities as well). Mothers’ statements indicated an understanding of how these practices help child development and why they are important.

These results align with the literature on supporting positive parenting. Positive parenting can promote the long-term health and well-being of children (Morris et al., 2017; Sandler, Ingram, Wolchik, Tein, & Winslow, 2015). For example, parental sensitivity in early childhood, as demonstrated in this study, can result in long-term positive child outcomes such as enhanced social skills and academic achievement (Raby, Roisman, Fraley, & Simpson, 2015). Likewise, parental involvement can lead to improvements in social skills and reductions in problem behaviors (El Nokali, Bachman, & Votruba-Drzal, 2010). Appropriate levels of parental control and demands can protect against behavior problems and externalizing disorders well into adolescence, as reviewed by Hoskins (2014). Finally, practices that stimulate children’s development, like reading together as discussed by mothers in this study, have demonstrated positive effects in vocabulary and comprehension as well as later language and cognition (Raikes et al., 2006).

These results also align with the prior findings on child outcomes from the Legacy RCTs, and offer preliminary support for the effects of the intervention activities on Legacy goals as described in the Legacy logic model (Figure 1). Taken alongside quantitative findings, this knowledge can inform future dissemination and scaling up of Legacy by helping to understand what Legacy mothers at each site value regarding their growth as parents, and what Legacy concepts and strategies they apply at home. Already, examination of the focus-group data has shaped the creation of intervention training and enrollment materials, and helped to develop implementation supports and technical assistance for current and future implementers of the Legacy program (Robinson, Perou, & Leeb, 2014). As a Spanish language adaptation of Legacy is being developed, a similar complementary qualitative approach is being used to ensure the cultural congruence and effectiveness of the Legacy program in a community-based implementation.

Finally, these findings contribute to methodological advancements in the evaluation of parenting programs overall. Recent research has emphasized the value of combining qualitative methods with quantitative ones—often within RCTs—to evaluate interventions (Catallo, Jack, Ciliska, & Macmillan, 2013; Furlong & McGilloway, 2012; O’Cathain et al., 2014; O’Cathain et al., 2013). Although some research has examined changes in parenting, many qualitative studies of parenting programs have focused more on areas such as participant satisfaction, and facilitators and barriers to engagement (Koerting et al., 2013; Rahmqvist, Wells, & Sarkadi, 2014), and have looked less at specific parenting practices. Finally, while prior studies have focused less on secondary analyses of qualitative data, there is evidence that this method can serve as an effective tool when aligned with the aims of the original study (Corti & Bishop, 2005; Heaton, 2000).

A notable finding of this study consists of the differences between sites related to characteristics and strategies of parenting described. In Miami, mothers described the necessity of making life changes prior to increasing commitment to parenting, such as reducing their amount of partying. These themes did not emerge in Los Angeles, suggesting that Los Angeles mothers may be focusing on different aspects of the parenting role. Mothers in Miami also made more varied statements about the practices discussed, including different and often contrasting views related to commitment and parental control. For example, some mothers in Miami discussed the challenges in using noncorporal techniques while mothers in Los Angeles did not. Despite these differences, discussion among mothers at both sites indicates that Legacy successfully encourages mothers to be reflective in their consideration of alternatives to spanking and changes in their parenting role. In Los Angeles, mothers also provided greater detail and more specific examples of their parenting cognitions and behaviors, as compared to mothers in Miami. In addition, mothers in Los Angeles described not only being sensitive to their children but also taking the time to explain behaviors and consequences to their children. In this respect, the discourse of mothers in Los Angeles demonstrated an increased capacity to discuss and interpret how their relationship with their child functions.

These site differences may not be as surprising considering the demographic differences across the sites. Although both Legacy sites recruited participants based on poverty criteria, the sites varied significantly on demographic variables such as maternal education, maternal IQ, employment status, marital status, and racial/ethnic composition, with the Los Angeles sample overall having more social and economic resources. These demographic factors have all been previously associated with parenting (Ellingsen et al., 2014; Fox, Platz, & Bentley, 1995; Lucas-Thompson, Goldberg, & Prause, 2010; Quintana et al., 2006). Previous research also has demonstrated a relationship between adult cognitive skills, such as working memory, and poverty-related chronic stressors (Evans & Schamberg, 2009). Alternatively, mothers in Los Angeles in general simply may have been better able to articulate changes in their parenting, as compared to mothers in Miami.

Differences across sites, taken in conjunction with the overarching positive parenting results among mothers in Miami and Los Angeles, have important implications for program developers and implementers. As evident in this study, mothers in different contexts and with different backgrounds may present varying needs and outcomes related to parenting—for example, in receiving and processing information about discipline. Legacy aims to meet mothers wherever they are in their development as parents, and recognizes the importance of fostering mothers’ sense of self-efficacy. This study indicates that even alongside curriculum and individual-level differences, Legacy promotes knowledge and use of positive parenting practices. Legacy implementers should continue to follow the Legacy approach of reflecting on and responding to each mother’s individual needs. In terms of larger program dissemination, these findings suggest the importance of creating a safe, nonjudgmental group environment where mothers can cognitively reflect on current and past parenting practices and try out new ones. This highlights the importance of training and staffing supports. Nonjudgmental and nondidactic implementation over the long-term necessitates special skills and experience; in-depth and consistent selection and training of group leaders may significantly affect mothers’ participation and engagement in the program, and the overall effect of Legacy on their parenting. In addition, this work, similar to other quantitative and qualitative research with low-resourced families (Landy, Jack, Wahoush, Sheehan, & MacMillan, 2012; National Center for Parent, Family and Community Engagement, 2015), supports the Legacy tenet that parents can make positive changes that support their children’s development even when they are experiencing the stressors, disadvantage, and chaos of a poverty environment. While additional research would clarify the generalizability of Legacy to other populations, these findings are supportive of dissemination of the program on a broader scale.