Engaging Fathers in Effective Parenting for Preschool Children Using Shared Book Reading: A Randomized Controlled Trial

Abstract

Engaging fathers and improving their parenting and, in turn, outcomes for their children in preventive/promotion-focused parenting interventions has been a notable, but understudied, challenge in the field. This study evaluated the effects of a novel intervention, Fathers Supporting Success in Preschoolers: A Community Parent Education Program, which focuses on integrating behavioral parent training with shared book reading (i.e., Dialogic Reading) using key conceptual models (i.e., common elements, deployment model, task shifting) to engage and improve father (i.e., male guardians) and child outcomes. One hundred twenty-six low-income, Spanish-speaking fathers and their children were recruited across three Head Start centers in urban communities and were randomized to the intervention or to a waitlist control condition. Outcomes were obtained before and immediately postintervention and included observed and father-reported parenting and child behaviors, standardized assessments of language, and father self-reported parental stress and depressive symptoms. Attendance data were also collected as a proxy measure of engagement to the intervention. Parenting behaviors (observed and father-reported), child behaviors (father-reported), and language development of the children in the intervention group improved significantly relative to those in the waitlist control condition. Effect sizes (ESs) were in the small to large range across outcomes. Fathers can be engaged in parenting interventions, resulting in improved parent and child outcomes. Greater attention must be given to methods for maximizing parenting within a family and toward developing effective, engaging, and sustainable intervention models for fathers.

Fathers play a significant role in the social, emotional, and behavioral development of children (Lamb & Lewis, 2013; Panter-Brick et al., 2014). For example, fathers who are actively involved in their children’s lives are more likely to have children with fewer disruptive behavior problems over time (Ramachandani et al., 2013). Moreover, fathers who are sensitive and attuned have children who are better adjusted psychosocially, even when taking into account the mothers’ parenting behavior (Grossman et al., 2002). Thus, fathers and male guardians represent an important group to target for behavioral parent training (BPT)—an intervention with robust empirical support documenting improvements in parenting behavior and child functioning (Fletcher, Freeman, & Matthey, 2011). Nevertheless, there is a paucity of empirical research specifically focused on improving the quantity and quality of fathers’ parenting behaviors, as most BPT research has often included only mothers as participants (Fabiano, 2007; Panter-Brick et al., 2014).

In the relatively small number of research studies that have investigated father outcomes, results are mixed. Father involvement in BPT can improve parenting behavior (e.g., Danforth, Harvey, Ulaszek, & McKee, 2006; Fabiano et al., 2012; Frank, Keown, & Sanders, 2015) and directly benefit children (Danforth et al., 2006; Frank et al., 2015; Lundahl, Tollefson, Risser, & Lovejoy, 2008), although benefits are not always evinced (see Fabiano, 2007; Panter-Brick et al., 2014, for reviews). In particular, efforts at utilizing BPT specifically for fathers in the context of prevention/promotion-focused efforts have been limited (see Panter-Brick et al., 2014). As an example, in a randomized controlled trial of BPT for fathers, Helfenbaum-Kun and Ortiz (2007) investigated the effects of a father-only BPT prevention group in a sample of low-income Head Start families on multiple father and child outcomes. Relative to a waitlist control group, fathers who were enrolled in the group BPT program did not improve on parenting or child outcomes, likely because 70% of fathers dropped out of the BPT group. This study aligns with others that suggest that BPT may be an effective intervention for fathers when actively implemented, but poor engagement often attenuates potential benefits (Fabiano, 2007; Panter-Brick et al., 2014).

One potential reason for the challenges in engaging fathers in BPT may be the content and structure of BPT interventions. The parenting literature for fathers focuses on a deficit model that depicts fathers as ineffective or neglectful parents (Fabiano, 2007; Panter-Brick et al., 2014). Research has documented that fathers report a desire to be good parents and to help their children achieve positive outcomes (Fabiano, Schatz, & Jerome, 2016), but barriers, such as less inclination to accept services and a sense that services do not appropriately address their needs, limit their success (Dubowitz, Lane, Ross, & Vaughan, 2004). As noted by Fabiano (2007), the content of typical BPT programs assume a parenting deficit that needs to be “fixed,” and this model may disrupt engagement and efforts toward change. Moreover, fathers serve in different roles in many families, and the topics discussed in BPT more often focus on the roles typically held by mothers (e.g., daily household activities). As such, a strength-based approach focused on meaningful father–child interactions that addresses important functional child outcomes, such as learning to read (Duursma & Pan, 2011; Fabiano et al., 2016; Fagan & Iglesias, 1999), may be helpful to engage fathers in BPT.

In addition, given the empirical evidence for BPT, as well as other evidence-based interventions, increased attention has been paid to whether these interventions can be delivered in routine service settings (e.g., schools, pediatric care, community organizations) and be sustainable over time (Axford, Lehtonen, Kaoukji, Tobin, & Berry, 2012; Bumbarger & Perkins, 2008; Weisz, Ugueto, Cheron, & Herren, 2013), particularly for fathers (Chronis, Chacko, Fabiano, Wymbs, & Pelham, 2004; Panter-Brick et al., 2014). As Panter-Brick and colleagues (2014) noted, it is important to focus on the effectiveness of an intervention while determining the extent to which the intervention can reach fathers and be sustainable in the proposed setting. Weisz and colleagues (2013) found that less than 2% of clinical trials utilize clinically referred youth being treated by service providers within the practice setting. The modal intervention trial is usually conducted within research/university lab settings with research staff (e.g., graduate students) as therapists, who are trained intensively and supervised by the developer or an expert in the intervention.

Although this discrepancy between research and practice settings is less prominent in prevention studies, there remain parallel issues (Axford et al., 2012; Bumbarger & Perkins, 2008). As an example, whereas prevention-based studies are often conducted in the practice/service setting, these studies often utilize highly trained and supervised research staff to implement the intervention (e.g., Baydar, Reid, & Webster-Stratton, 2003; Brotman et al., 2005). Collectively, it is not surprising that the effects of these preventive interventions are often less robust when implemented within routine practice constraints (Bumbarger & Perkins, 2008). Clearly, there is a need to consider the broad context, constraints, and key priorities of the practice setting, given an end goal of intervention research should be dissemination to these practice settings. Development of interventions that take into account the skill level of intervention providers/supervisors, training and implementation demands, the financial constraints of the practice setting, the interests of the target population, and the priorities of the practice setting are more likely to result in an intervention that balances access, engagement, effectiveness, reach, and sustainability. Striking this balance is likely necessary to maximize the potential of parenting interventions delivered in community settings.

The development of the Fathers Supporting Success in Preschoolers: A Community Parent Education Program (FSSP program) is an effort to advance the field of parenting interventions for fathers. The FSSP program is a preventive/promotion-focused intervention developed by community partners (i.e., Head Start staff and a father community advisory board) in collaboration with researchers to improve positive parenting behaviors of fathers (i.e., male guardians with a consistent parenting role). The program was designed to target parents of young children enrolled in Head Start in low-income, urban communities. As more fully described in the Methods section, we utilized the deployment model (Weisz, Jensen, & McLeod, 2005) as a framework, as we have done in previous projects (Chacko et al., 2015), to develop an intervention while “working with end in mind” (i.e., an intervention that is accessible, engaging, effective, and sustainable in Head Start). In development of the intervention, we utilized evidence-based common elements of BPT (e.g., praise, effective communication; Chorpita & Daleiden, 2009; Garland, Hawley, Brookman-Frazee, & Hurlburt, 2008; Kaminski, Valle, Filene, & Boyle, 2008). As we have discussed in previous efforts (Chacko et al., 2015), a common elements approach may offer significant advantages for providing effective, engaging, and sustainable interventions. Further, methods that are established for deploying parenting training in groups were integrated into the development of the program (Cunningham, Bremner, & Secord-Gilbert, 1998; Whitehurst, Arnold, et al., 1994).

Focusing on a select number of common element content and methods to teach BPT allows for facilitators, particularly those without significant training in BPT, to learn the content and methods more specifically and intensely, as well as to apply these with greater fidelity. From a father’s perspective, focusing on a select number of BPT skills allows one to implement those skills more efficiently and through methods that are likely to result in skill acquisition. From the perspective of Head Start administrators, focusing on common elements and methods may result in easier training, implementation, and supervision. Collectively, leveraging common and effective content and methods allows for a more streamlined and efficient intervention.

The FSSP program delivery occurred in a group-based, highly interactive, father-to-father local knowledge transmission format (i.e., coping–modeling–problem-solving group; Cunningham et al., 1998). The coping–modeling–problem-solving subgroup approach has been successfully used to engage parents in several BPT studies (Chacko, Wymbs, Chimiklis, Wymbs, & Pelham, 2012; Chacko et al., 2008, 2009; Cunningham et al., 2000; Fabiano et al., 2012). The coping–modeling–problem-solving subgroup approach pioneered by Cunningham (Cunningham et al., 1998) utilizes videotaped vignettes of exaggerated errors made by parents in common parent–child interactions to generate discussion and learning. After viewing a vignette in the large group, smaller groups of parents (i.e., subgroups) identify the errors in parenting and suggest alternative approaches. In the larger group, subgroups share their answers and group facilitators utilize attribution questions (e.g., What is the consequence of this approach? What message does this approach convey to the child? What is the short- and long-term impact of this approach on the child, the father–child relationship, and the family?) to help support deeper conversation and understanding of the effects of parenting errors and the alternative parenting strategies. This approach to BPT has been effectively used to support the presentation of BPT principles to families in large groups (e.g., Cunningham et al., 2000).

Videotaped modeling of parenting has been incorporated into a range of BPT programs to provide parents with the opportunity to gain exposure to new strategies, increase engagement, and generate discussion of key principles and strategy use (Chacko et al., 2009; Cunningham et al., 2000; Fabiano et al., 2012; Gross, Garvey, Julion, & Fogg, 2007; Webster-Stratton, 1990). Based on social learning theory, existing evidence suggests that videotaped modeling is an effective method for learning behavioral principles and parenting techniques (Breitenstein, Gross, & Christophersen, 2014; O’Dell et al., 1982). Studies suggest that viewing videotaped vignettes without coaching can result in significant changes in parenting and child behavior for families with young children with conduct problems (Webster-Stratton, 1990) and can even lead to the most consistent acquisition of parenting skill compared to other methods (written information, audiotaped information; live modeling and coaching; O’Dell et al., 1982). The process of analyzing parenting errors and suggesting alternative approaches might be particularly helpful for fathers in that they are encouraged to apply their own expertise in discussion of parenting versus creating an implicit sense of critique by learning these skills directly from expert leaders (Fabiano et al., 2012). Videotaped modeling seems to be most effective when the videotapes depict families in pertinent contexts who seem similar to the target population in racial and ethnic background (Gross et al., 2007).

The novel aspect of the FSSP program is that we integrated BPT skills training into a key parent–child context (i.e., shared book reading) using principles of Dialogic Reading (DR), an evidence-based intervention to improve early academic readiness (primarily language skills) in preschool children (Lonigan & Whitehurst, 1998; Wasik & Bond, 2001; Whitehurst, Arnold, et al., 1994; Whitehurst, Epstein, et al., 1994; Whitehurst et al., 1999). Shared book reading through DR was espoused by Head Start, the father’s community advisory board, and researchers as a meaningful context for supporting father–child interactions and improving parenting behavior (see also Fabiano et al., 2016) while also aligning closely with one of the key priorities of early education programs (i.e., developing academic readiness). In addition, fathers often engage in this activity with their children (Duursma & Pan, 2011) and both mothers and fathers value and hope to support their young children’s learning (Duursma & Pan, 2011; Fabiano et al., 2016; Fagan & Iglesias, 1999). Moreover, researchers have successfully recruited fathers of preschoolers for DR (e.g., Fagan & Iglesias, 1999), including those from lower socioeconomic status backgrounds (Whitehurst, Arnold, et al., 1994). The approach of FSSP is novel; at the present time, there has not been a study to investigate the efficacy of integrating BPT with DR for fathers of young children. In one previous study, researchers endeavored to examine the independent and concurrent effects of BPT and DR for preschool children (Arnold et al., 2006), but problems with attrition limited power and precluded examination of intervention effects (D. Arnold, personal communication, October 3, 2016).

Studying BPT combined with shared book reading may hold merit because rather than a global goal to “increase father involvement,” which implies a deficit approach, it targets a specific parenting skill-set and represents a valued activity (Lareau, 2000), an approach used successfully for engaging fathers of older children (Fabiano et al., 2012). By focusing on DR, fathers may be more receptive to learning and practicing parenting skills. Although playing games with children may feel culturally dissonant for some fathers (Gross et al., 2007), sharing a book together is an activity that promotes joint exploration without competition. Using DR as a vehicle to practice parenting skills may allow fathers to support the development of child learning competencies while practicing parenting skills to support development. Ultimately, we believe that developing a program that is not explicitly parent focused but parent and child focused, one that is not deficit driven (improving problematic parenting) but focused on skill development (improving children’s language skills), one centered on identified areas of interest of/relevance to fathers (academic readiness through shared book reading) rather than those which may be less relevant (getting the child to do chores) and developmentally appropriate for the child would be an intuitively appealing and engaging format for fathers to learn how to enhance their parenting.

Our goal in the FSSP program was to determine if key targets of BPT (e.g., parenting and child behavior) and DR (language) would improve following the integration of these interventions within the context of the FSSP program. Last, given the role of parenting stress and depressive symptoms in the lives of low-income families (Chronis et al., 2004), we aimed to determine the effects of BPT on these outcomes. The purpose of this study was to evaluate the effects of the FSSP program on improving key proximal outcomes (i.e., parenting), secondary outcomes (i.e., child behavior and language), and distal outcomes (i.e., parental stress and depressive symptoms), compared to a waitlist control group, through a randomized controlled trial design. It was hypothesized that relative to those in the waitlist control group, fathers randomly assigned to the FSSP program would demonstrate significant improvements in proximal outcomes of parenting behavior (self-report and observed parenting), secondary outcomes of child behavior (parent report and observation) and child language (on standardized language assessments), and distal targets of parenting stress and depressive symptoms (parent report).

METHOD

Participants

Recruitment for the study occurred between September 2010 and February 2013 across three Head Start centers in New York City. Inclusion criteria included the following: (a) the child had to attend one of the participating Head Start centers, (b) the father (i.e., a male guardian with a consistent parenting role [e.g., biological father, stepfather, grandfather, uncle, mothers’ significant other]) had to have consistent involvement in the caregiving of the target child, and (c) the father and child had to speak English and/or Spanish. A total of 126 father–child dyads were enrolled (see Table 1 for demographic information). The sample primarily consisted of Latino, Spanish-speaking biological fathers who were married to the child’s biological mother. The sample consisted primarily of low-income families based on federal eligibility guidelines for Head Start. The majority of fathers reported not completing high school.

TABLE 1.

Demographics of Study Participants by Group

	FSSP (n = 64)	Waitlist (n = 62)
Child Age in Years	M = 4.76 (SD = .32)	M = 4.42 (SD = .36)
Child Sex	66% male	69% male
Father Age (Years)	M = 36.77 (SD = 8.56)	M = 35.25 (SD = 8.65)
Father Education (Years)	M = 8.28 (SD = 1.95)	M = 8.22 (SD = 1.88)
Father Race/Ethnicity	85% Hispanic/Latino	89% Hispanic/Latino
	78% White	82% White
	17% Black/African American	14% Black/African American
	13% Did not report ethnicity	11% Did not report ethnicity
	5% Did not report race	4% Did not report race
Income (% Less Than $24,000)	70%	68%
Father Primary Spoken Language	89% (Spanish)	90% (Spanish)
	11% (English)	10% (English)
Father Marital Status (% Married)	62%	59%

Note There were no significant differences between groups on any variables in the table. Some percentages sum to greater than 100% due to rounding error. FSSP = Fathers Supporting Success in Preschoolers: A Community Parent Education Program.

Procedures

The Institutional Review Board approved the study. After a complete description of the study was provided to the participants, written informed consent was obtained from the father. Assent was not required, per the Institutional Review Board, given the age of the participating child. In participating Head Start centers, research staff and Head Start staff conducted informational sessions about the study with families, distributed flyers to each family, and used meet-and-greet opportunities at child arrivals and departures for recruitment. Interested families were scheduled to complete consenting and completion of all assessments and observations with research staff at the Head Start center. Participants were randomly assigned to treatment condition (FSSP = 64, waitlist = 62; see Figure 1) by senior research staff (blind to participant profile) based on a random permutation calculator (https://www.randomizer.org/). This article focuses on assessments at baseline and at completion of the program in order to determine immediate benefits of the FSSP program.

FIGURE 1.

Consort diagram of participants in the study. FSSP = Fathers Supporting Success in Preschoolers: A Community Parent Education Program.

Intervention Condition

The FSSP program is an 8-week, 90-min-per-week, group-based intervention held during the late afternoon hours at the child’s Head Start center. Within the context of shared book reading, DR forms the foundation of the FSSP program, with common elements of BPT integrated and generalized through a coping–modeling–problem-solving subgroup approach. Videotaped vignettes of father–child reading interactions were used to provide opportunities for small- and large-group discussions. Groups ranged in size from six to 12 fathers, with small subgroups consisting of three or four fathers. Fathers had access to age-appropriate books through the participating Head Start site to complete homework and practice implementation of DR at home. All materials, including the FSSP facilitator manual, parent manual, and videotaped vignettes, were developed in both English and Spanish. Each FSSP session included a meal for families, but no other incentives were provided for participation in the FSSP program. Although parents participated in group sessions, children (target child) engaged in child-focused activities (e.g., arts and crafts) led by one or two adult facilitators.

Shared Book Reading Through DR

In traditional reading, the parent is the storyteller and the child the passive listener; however, in DR, the parent uses prompts and feedback sequences to allow the child to become an active storyteller (Whitehurst, Arnold, et al., 1994). The process relies heavily on the use of pictures, so the father does not have to be literate. Feedback includes ample praise and encouragement for the child’s efforts, correction without criticism, and tailoring of feedback and questions to the child’s developmental level, all of which are also components of effective BPT approaches. Additional guiding principles include following the child’s lead, being sensitive to the child’s interests, varying the types of questions, and keeping the reading interaction fun.

Integration of Common Content and Methods of BPT

In order to target and ultimately improve outcomes, we integrated key, common BPT content and methods within DR. DR was used as a vehicle to deliver BPT content and methods in an indirect, child-focused context. Ultimately, the goal was for these BPT principles to be generalized from the DR context to other areas of father–child interaction.

As has been done in our previous intervention development efforts (Chacko et al., 2015), as well as those specifically for fathers (Fabiano et al., 2012), we utilized a common elements approach (e.g., Chorpita & Daleiden, 2009; Garland et al., 2008; Kaminski et al., 2008) by integrating essential content and methods from BPT into DR. Core content of BPT integrated into FSSP include those that focus on antecedent behaviors (e.g., effective commands) and consequences (e.g., labeled praise). Content was taught/delivered through common methods utilized in BPT programs (e.g., discussion, role-play, in-session father–child practice of skills with feedback, homework). Specifically, sessions focused on the following BPT content integrated with shared reading: (a) improving parent–child relations through child-centered time wherein fathers could follow the child’s lead, (b) establishing consistent routines, (c) using attention to promote positive behaviors, (d) using ignoring and distraction to reduce attention-seeking behaviors, (e) providing clear and developmentally appropriate expectations, (f) utilizing incentives for positive behavior, (g) helping children learn consequences of their behavior, (h) providing consistent expectations, and (i) using time-out sparingly for challenging behavior.

Father-To-Father (Local Knowledge) Facilitation Through Subgroups

Within the large group of the FSSP program, fathers view videotapes of exaggerated parenting errors that occur in the context of shared book reading. Fathers then break into small subgroups, identify the errors of the fathers in the videos, and brainstorm strategies that the fathers in the videos could have used to be more successful. Moreover, facilitators utilize attribution questions (e.g., what messages does this approach convey to the child?) to help support learning and motivation to implement parent-derived solutions. Important to note, although the majority of the FSSP session is focused on evidence-based parenting approaches in the context of shared-book reading, a significant segment of each session is focused on having fathers discuss if, when, and how these parenting approaches can be utilized in other father–child contexts through the same coping–modeling–problem-solving subgroup process. As such, fathers are supported to generalize self-identified and ecologically valid but evidence-based parenting behaviors, to many relevant parenting contexts, and to discuss these within their subgroup and with the larger group. The generalization of fathers’ self-identified and ecologically valid parenting skills is further supported through weekly homework assignments.

Measures

All baseline and posttreatment measures were collected at the participating Head Start center. Fathers were compensated $20 for completion of baseline assessments and $40 for completion of posttreatment assessments.

Attendance

Fathers’ attendance at each FSSP session was coded as present or absent. The percentage of sessions wherein fathers attended FSSP was utilized as a measure of engagement.

Parent Behavior Checklist

The Parent Behavior Checklist (Fox, 1994) is a parent-report measure used to assess parenting strengths and weaknesses. The scale consists of 100 items and includes three empirically derived subscales: Expectations (parent’s developmental expectations for child), Discipline (parent’s coercive/harsh responses to difficult child behavior), and Nurturing (parent’s promotion of the child’s psychological growth). Reported alpha and test–retest reliability coefficients, respectively, are .97 and .98 for the Expectations subscale, .91 and .87 for the Discipline subscale, and .82 and .81 for the Nurturing subscale. In the current study, Cronbach’s alpha was .90, .88, and .82 for the Expectations, Discipline, and Nurturing subscales, respectively. Normalized T scores are reported in Table 2. T scores between 1 and 30 are classified as lower extreme, 31 and 40 as significantly below average, 41 and 59 as average, 60 and 69 as significantly above average, and at or above 70 as upper extreme.

TABLE 2.

Means, Standard Deviations, and Effect Size Data for Dependent Variables for Treatment Groups

	FSSPa		Waitlistb
Variable	Baseline M (SD)	Post M (SD)	Baseline M (SD)	Post M (SD)	ES (d)
PBC–Expectations	45.89 (12.34)	50.73 (13.34)	44.32 (13.21)	49.11 (13.67)	.12
PBC–Discipline*	61.56 (8.23)	51.32 (8.10)	60.01 (9.29)	58.45 (9.33)	.82
PBC–Nurturing*	45.11 (7.23)	54.97 (8.33)	46.77 (8.65)	47.11 (8.99)	.91
PSI–SF	73.5 (22.76)	61.31 (22.12)	69.94 (23.12)	60.45 (21.16)	.04
CESD	9.22 (2.10)	8.75 (2.46)	9.32 (2.35)	8.98 (2.21)	.10
DPICS–Positive Parenting*	12.54 (13.89)	22.89 (16.32)	12.33 (12.01)	13.21 (14.23)	.63
DPICS–Negative Parenting*	14.23 (14.67)	8.33 (10.11)	14.01 (14.22)	14.98 (14.56)	.53
DPICS–Child Problems	16.78 (16.71)	10.43 (11.23)	17.01 (16.89)	15.23 (16.54)	.34
ECBI–Intensity*	89. 56 (31.25)	72.21 (28.22)	91.32 (30.25)	87.01 (29.26)	.51
PLS–Auditory*	80.45 (11.56)	88.44 (12.22)	78.14 (12.01)	82.12 (11.89)	.52
PLS–Expressive*	81.33 (11.78)	89.89 (12.23)	80.21 (12.12)	83.98 (11.01)	.51

Note FSSP = Fathers Supporting Success in Preschoolers: A Community Parent Education Program; Post = posttreatment; ES (d) = effect size (Cohen’s d) calculation comparing FSSP group to waitlist control group (positive effect sizes represent improvement due to the FSSP program); PBC = Parent Behavior Checklist; PSI-SF = Parenting Stress Index–Short Form; CESD = Center for Epidemiologic Studies Depression Scale; DPICS = Dyadic Parent–Child Interaction Coding System; ECBI = Eyberg Child Behavior Inventory; PLS = Preschool Language Scales.

^an = 64.

^bn = 62.

^*Significantly different from mean of waitlist control group at posttreatment (p < .01).

Dyadic Parent–Child Interaction Coding System

Observed parenting and child behavior was measured using the Dyadic Parent–Child Interaction Coding System-R (DPICS-R; Robinson & Eyberg, 1981). The validity of the DPICS-R has been evaluated in multiple studies, including those in Head Start (Webster-Stratton, Reid, & Hammond, 2001). For the purposes of this study, the composite scores for Positive Parenting (praise, positive affect, and physical positive), Negative Parenting (critical statements), and Child Problems (whine, cry, negative physical behavior, smart talk, yell, destructive, and noncompliance) were used. A team of two independent coders were trained and coded all of the tapes, with 40% of these tapes double-coded to assess interrater reliability. Interobserver agreement coefficients were .81 for Positive Parenting, .82 for Negative Parenting, and .81 for Child Problems.

Three standard parent–child interactions that were 5 min each were coded using the DPICS-R. In the first 5-min segment, the father was instructed to allow the child to lead the play (i.e., father plays with whatever child chooses and according to the child’s rules). During the second 5-min segment, the father was instructed to lead the play (i.e., father chooses the play activity and sets the rules). During the last 5-min segment, the father was instructed to have his child clean up the toys independently.

Eyberg Child Behavior Inventory

The Eyberg Child Behavior Inventory (ECBI; Eyberg & Pincus, 1999) is a 36-item parent rating scale of a child’s disruptive behavior measured across two scales. The Intensity Scale measures the frequency of disruptive behavior on a scale from 1 (never) to 7 (always). An Intensity score above 131 is considered clinically significant. Although not used in the present study, the Problem Scale asks whether each behavior is a problem for the parent on a yes–no scale. The Intensity and Problem scales have high internal consistency coefficients (Cronbach’s α = .97 and .95, respectively), as well as good stability coefficients across a 12-week interval (.80 and .85, respectively; Eyberg&Pincus, 1999). The ECBI has been used to assess intervention effects in numerous studies, including parenting interventions for preschool children in Head Start (Webster-Stratton et al., 2001). In this sample, Cronbach’s alpha was .89 for the Intensity Score.

Preschool Language Scales (Fourth Edition)

The Preschool Language Scales (4th edition; PLS-4; Zimmerman, Steiner, & Pond, 2002) is an individually administered test of language for children from birth through 6 years 11 months. The PLS-4 targets receptive and expressive language skills in the areas of attention, play, gesture, vocal development, social communication, vocabulary, concepts, language structure, integrative language skills, and phonological awareness. The PLS-4 has two subscales. The Auditory Comprehension subscale is used to evaluate how much language a child understands (receptive language), whereas the Expressive Communication subscale is used to determine how well a child communicates with others (expressive language). Receptive and expressive language have been shown to improve following DR (Wasik & Bond, 2001). The PLS-4 was used to confirm the effects of DR as applied in the FSSP program. A Spanish version of the PLS-4 was used when the child’s teacher stated that Spanish was the child’s preferred language. The Spanish version of the PLS-4 was used with 40 (32%) of the children in the study. Extensive psychometric data is available on the English and Spanish version of the PLS-4 (see PLS-4 technical manual; Zimmerman et al., 2002). Standard scores for both subscales were used in the analyses.

Parenting Stress Index-Short Form

The Parenting Stress Index–Short Form (Abidin, 1995) is a 36-item self-report measure used to assess parental stress in three domains (i.e., parental distress, parent–child dysfunctional interaction, and difficult child) and includes a total score. The total score was used in the analyses. Responses on the Parenting Stress Index–Short Form are measured along a 5-point scale ranging from 1 (strongly disagree) to 5 (strongly agree), with higher total scores indicating greater levels of parental stress. Total scores range from 36 to 180, with scores above 90 indicating clinically significant total stress. High levels of internal consistency (α = .83) and 1-year stability (r = .75) have been reported. In the current study, Cronbach’s alpha was .87 for the total score.

Center for Epidemiologic Studies Depression Scale

The Center for Epidemiologic Studies Depression Scale (Radloff, 1977) is a screening tool consisting of 20 items that assess the frequency of depressive symptoms occurring in the past week (0 = rarely/none of the time, 1 = some or a little of the time, 2 = occasionally or a moderate amount of time, 3 = most or all of the time). Center for Epidemiologic Studies Depression Scale scores are summed and range from 0–60, with a score of 16 and above considered clinically significant (Zich, Attkisson, & Greenfield, 1990). The current study had good internal consistency with a Cronbach’s alpha of .90.

Therapists, Training, Treatment Fidelity, and Supervision

Six FSSP groups were conducted across three Head Start sites with a total of seven parent-group cofacilitators and nine child-group facilitators. Facilitators were selected by Head Start administrators or volunteered to participate in the study. Parent-group facilitators were diverse, with one teacher, two assistant teachers, two front-desk receptionists, and two social workers. Child-group facilitators were assistant teachers at the Head Start site.

Initial training was provided by the FSSP developers to all FSSP parent facilitators during three 2-hr in-person trainings that occurred weekly for 3 consecutive weeks. The focus of the training was on the content and process of the FSSP, as well as understanding and implementing DR and common elements of BPT. Training methods included reading of the FSSP manual, didactic lecture, modeling, roleplay, and completion of homework assignments. Based on posttraining evaluations, which included a written test and role-plays of key processes within FSSP, further training was conducted for two of the seven FSSP parent facilitators who did not attain competency after the initial training. These two facilitators were provided additional training (i.e., additional 2 hr of training) to achieve initial levels of competence. Child-group facilitators completed a 1-hr training focused on group-management procedures.

Fidelity monitoring tools were used to assess parent facilitator adherence to curriculum content (e.g., session topics, information conveyed, competence in implementing key methods) and the process of facilitation (e.g., information exchange, competence in guiding discussions, use of COPE attributional questions). All FSSP sessions were videotaped, and these fidelity assessments were used by research staff to assess for treatment fidelity. Fidelity data demonstrated that FSSP facilitators implemented the FSSP intervention with an average rate of fidelity equaling 88%. In addition, senior research staff held brief (30–45 min) weekly group supervision with each facilitator team to review key issues gleaned from the videotaped fidelity assessments, discuss content and process of the groups, and discuss challenges in implementation or issues that arose during the group sessions. Videotaped FSSP sessions also served as opportunities for additional adaptive training. Specifically, additional training was provided when fidelity ratings for FSSP facilitator pairs fell below 80% for two consecutive FSSP sessions. Additional training was adapted (i.e., individualized) based on area of need (e.g., support in utilizing COPE attribution questions, maximizing utility of vignettes). Additional remedial training was provided for two FSSP facilitator pairs across the duration of the study due to low levels of adherence (< 80%). Fidelity was not conducted for the child groups.

Data Analysis

Random-effects regression (RR) using SAS PROC MIXED procedure (SAS v. 9.4) was used for primary intent-to-treat analyses for the specific aims of this study. All baseline demographics and outcome measures were statistically equivalent between intervention groups. Complete data were collected on 98% of participants at posttreatment. Last observation carried forward was utilized for missing data. Two sets of contrasts were tested for the specific aims of this study. First, analyses compared the FSSP program to the waitlist group. For each dependent variable, the first-order effect of time of assessment (i.e., baseline to posttreatment) was assessed, as well as first-order effects of each contrast (i.e., FSSP vs waitlist). Group × Time interactions were examined to determine if improvement in functioning over time was differentially associated with treatment group. Analyses controlled for attendance to FSSP. A conservative alpha level (i.e., p < .01) was adopted for these analyses given multiple significance tests. Last, ES estimates (Cohen’s d) were computed for each outcome measure.

RESULTS

Attendance

On average, fathers attended 79% (SD = 8.32%) of the eight FSSP sessions. Approximately 8% of fathers did not attend any FSSP sessions. Among those who did attend at least one session, approximately 89% attended at least four of the eight sessions, 77% attended at least six sessions, and 52% attended every session.

RR Analyses

Acute benefits of the FSSP program compared to the waitlist control group were observed on several outcomes (see Table 2). RR analyses indicate significant benefits of the FSSP on proximal outcomes of father report of discipline, t (204) = 4.49, p < .0001; nurturance, t(204) = 4.99, p < .0001; and positive, t(200) = 2.94, p = .0006; and negative, t(200) = 2.91, p = .0006, observed parenting. The FSSP program did not result in significantly better outcomes on father reported expectations for their child. On secondary outcomes, significant benefits of the FSSP program were found for father-reported intensity of child behavioral difficulties on the ECBI, t(117) = 2.89, p = .0035. In addition, the FSSP program resulted in improved Auditory Comprehension, t(200) = 2.74, p = .0009, and Expressive Communication, t(200) = 2.79, p = .0009, relative to the waitlist control condition. There was a trend toward statistically significant improvement in observed child behavioral difficulties, t(200) = 1.51, p = .062, with the FSSP program. Finally, on distal outcomes, the FSSP program did not result in statistically significant improvements in father-reported parental stress or depressive symptoms.

Effect Size

ES calculations were conducted for each outcome measure based on scores in the two groups. ES ranged from no effect to a large effect (.04–.91). ES for proximal targets of parenting were large for father-reported positive parenting (d = .91) and negative coercive parenting (d = .82), with moderate ES for observed positive (d = .63) and observed negative (d = .53) parenting. ES for secondary outcomes of child behavior and language ranged from a small effect for observed child problem behavior (d = .34) to moderate effects for father-reported child behavior (d = .51), auditory comprehension (d = .52), and expressive language (d = .51). ES for distal targets of parental stress (d = .04) and depressive symptoms (d = .10) were negligible (see Table 2).

DISCUSSION

Findings from the present study are consistent with research showing that BPT-focused interventions are effective in improving both parenting (observed and reported) and child behavior in young children from high-risk communities (Gross et al., 2009; Hurlburt, Nguyen, Reid, Webster-Stratton, & Zhang, 2013; Reid, Webster-Stratton, & Beauchaine, 2001). Important to note, the current study is one of the few studies of a BPT intervention focused on fathers using a randomized controlled trial and utilizing multiple methods of assessment (e.g., parent-report, direct observation, child assessments; Panter-Brick et al., 2014). This study aligns with others that suggest that BPT interventions can improve fathers’ parenting behaviors (Danforth et al., 2006; Fabiano et al., 2012; Fletcher et al., 2011; Frank et al., 2015) and directly benefit children (Danforth et al., 2006; Frank et al., 2015; Lundahl et al., 2008). More specifically, results of this study suggest that fathers reported improved discipline approaches and improved promotion of their children’s psychological growth. Moreover, observation of father–child interactions demonstrated that fathers in the FSSP program had lower rates of critical statements to their children and increased use of positive parenting behaviors (e.g., praise, affection). Given the importance of father involvement and parenting on child outcomes (Lamb & Lewis, 2013; Panter-Brick et al., 2014), the findings of this study are notable and further suggest that fathers’ parenting behaviors can have an impact on improving key child outcomes. In addition, it is significant that these findings were achieved in primarily immigrant, Spanish-speaking families from high-risk communities who were not seeking services for their child. Moreover, effects were achieved with a short-term, highly engaging intervention that was administered by existing staff at the Head Start centers, who were trained and supervised in an efficient manner, increasing the likelihood of feasibility and sustainability in Head Start.

Unlike other BPT studies, the FSSP program focused indirectly on parenting behaviors by using a meaningful, strengths-based context (i.e., shared book reading) that involved fathers and their children. Fathers were not recruited to work on parenting or reduce child behavior problems, but to learn, with other fathers, skills to support their children’s school readiness, which may remove stigma and support engagement and openness to considering how to best support development. This approach may have led to the high rates of engagement in the FSSP intervention (i.e., 79% average attendance rate), which is in contrast to those from other prevention programs (e.g., Baker, Arnold, & Meagher, 2011; Helfenbaum-Kun & Ortiz, 2007). From our perspective, common, evidence-based elements of parenting are a set of specific principles and skills that are evinced in any parent–child interaction. As such, all parent– child interactions provide opportunities to focus on improving parenting behavior. For fathers from high-risk populations who are not explicitly seeking services, such as those in this study, identifying a meaningful, engaging father–child context that serves as a platform for fathers to learn parenting skills in a systematic manner is essential. In this regard, shared book reading for fathers of young children appears to be an important parent–child context with enough structure to support systematic learning, at least for some types of families (see the Limitations and Future Directions section regarding exceptions).

A common elements approach may also offer significant advantages to providing effective, engaging, and sustainable interventions. As a function of utilizing common content and methods, training of facilitators and collaborating with fathers was streamlined and focused on a low number of skills. This approach resulted in easier training, implementation (by facilitators and fathers), and supervision. Collectively, leveraging common and effective content and methods allows for a more streamlined and efficient intervention. Ultimately, this relative simplicity may allow for greater longer term sustainability of the FSSP program within Head Start.

The coping–modeling–problem-solving subgroup approach (Cunningham et al., 1998) may also offer several benefits to improving engagement, implementation, adherence, and sustainable practice. First, the subgroup approach allows fathers an opportunity to realize the consequences of parenting behaviors. This active approach also yields solutions with high levels of ecological validity in that fathers are collaborating on identifying the errors, the consequences of these errors, alternative approaches, and the consequence of the alternative approaches. In addition, because fathers are publicly stating solutions to the problems, commitment to change and feelings of competence are promoted. Finally, the approach may reduce resistance on the part of fathers. Given findings suggesting that men are reluctant to seek treatment for a problem central to their own self-concept (i.e., being a father; Addis & Mahalik, 2003), a parenting program that places men in the role of supporting their child’s development and as an “expert” in solving problems is promising. Moreover, data from our previous BPT study (Chacko et al., 2012) found that, relative to parents randomly assigned to a BPT collaborative group, those assigned to a BPT coping–modeling–problem-solving subgroup had greater treatment attendance and homework completion; rated their groups as being more socially supportive, with fewer barriers to treatment participation; and were less likely to drop out of treatment. In addition, given the focus on parent-to-parent facilitation and identification of ecologically valid methods to implement parenting practices through subgroups, facilitators are often less involved in providing input and suggestions during the session (Chacko et al., 2012). As such, facilitators truly facilitate the group and may require less knowledge about BPT content, methods, and processes. The implication of this is that potential intervention facilitators do not need substantial experience or expertise in BPT and that training, as well as supervision requirements, can be substantially reduced relative to traditional models where the group facilitator often leads a group. The benefit of the coping–modeling–problem-solving subgroup approach is particularly noteworthy in the context of Head Start, where staff often have limited experience in BPT. Ultimately, the coping–modeling–problem-solving subgroup approach may lead to increased sustainability.

This study adds to the literature on DR’s effects on emergent language capacity in preschool children from high-risk communities (Lonigan, Anthony, Bloomfield, Dyer, & Samwel, 1999; Wasik & Bond, 2001). It is important to note that the effects of FSSP (which integrates DR and BPT) must be put into context relative to the waitlist group. Although the waitlist group did not receive intervention as part of the research study, all children were participating and receiving academic readiness programs in Head Start. As such, the effects of the FSSP are above and beyond what is expected from Head Start alone. Given that there were moderate effect sizes on language outcomes, this suggests an educationally meaningful effect of the FSSP program, which adapted DR to incorporate BPT. This effect is not surprising given that DR is an established intervention, but effects on language were achieved even with this adaptation, speaking to the importance of integrating academic and behavioral programs to support multiple domains of development. These findings are consistent with research supporting the connection between social-emotional development and early academic skills (Arnold, Kupersmidt, Voegler-Lee, & Marshall, 2012). Of interest, the implementation of DR in this study may have been greater than what is typically employed. Some studies of DR have effectively trained parents in two 20- to 30-min sessions (Whitehurst et al., 1988) or using a 15-min instructional video (Blom-Hoffman, O’Neil-Pirozzi, Volpe, Cutting, & Bissinger, 2007). In the FSSP program, DR was the focus of eight consecutive weekly sessions where fathers learned the skills and then had opportunities for 10–15 min of direct implementation and feedback within the session. Daily homework activities were also assigned to implement DR and the parenting skills. Although we were unable to systematically collect data on homework completion, data from the supervision videotapes of the session suggest high levels of implementation of shared book reading at home. Collectively, the intensity of DR may have been greater than what is typically seen, and the resultant impact on language outcomes is not surprising.

Perhaps equally important, the focus of DR is on improving language skills and early academic readiness, which is well aligned with the standards of Head Start. Developing and delivering interventions consistent with the established standards of a practice setting are more likely to contribute to improved implementation and sustainability. Moreover, DR is a brief intervention in which Head Start staff can be readily trained to effectively implement without extensive training and supervision (Lonigan & Whitehurst, 1998).

There has been a growing recognition of the need to consider nontraditional professionals (e.g., Warner et al., 2016) as well as nonprofessionals (e.g., Calzada et al., 2005) in delivering interventions. Consideration must be given to expanding the potential workforce to enable dissemination and to have models from within the community who are advocates and teachers. This study adds to the literature on task-shifting (Lewin, Glenton, Gulmezoglu, Lavis,&Alvarez, 2012), which focuses on the use of nonprofessionals in the delivery of interventions, primarily in low resourced areas where access to evidence-based interventions are often limited. Inner-city communities, where the current study was implemented, represent the type of setting where task-shifting approaches are warranted. In Head Start study sites, an exclusive focus on delivering the FSSP program by more trained professionals (e.g., social workers) would have resulted in sites being unable to participate in the study and may have posed significant challenges with sustainability. Our approach of streamlining and simplifying the development of the FSSP program aligns with the task-shifting framework that allows for interventions to be more readily implemented by nonprofessionals (e.g., such as by a Head Start front-desk receptionist). Expanding the scope of the potential workforce to deliver interventions is important for effectiveness, and approaches that focus on utilizing key frameworks (e.g., task shifting, deployment models) may be particularly important in low-resourced communities where the demand may be great but the ability to deliver interventions is more limited.

LIMITATIONS

There are several notable limitations to this study. First, the majority of the Head Start sites that were approached to participate in the study, even though they expressed initial interest in participating, could not garner sufficient interest from fathers to enroll in a randomized controlled trial of the FSSP program. There likely were differences in programs (e.g., father involvement, organizational quality) that contributed to the ease with which fathers were recruited. This speaks to the issue of engaging fathers in interventions in general and further highlights the fact that the focus on reading may not be appealing for all fathers. A further limitation is that the present study provides information only about effects during the course of the intervention, so replication and follow-up data are needed to explore maintenance of treatment gains. The majority of fathers in this study were married/cohabitating, and generalization of engagement and outcomes must be evaluated with single/separated fathers. Moreover, significant benefits were not observed in all outcomes. Fathers did not report improvements in parental stress or depressive symptoms. This may not be entirely surprising, as the FSSP program did not specifically target these factors. Further, the baseline scores for fathers in these areas were within the normal range, suggesting that improvement in these areas was not indicated or expected. It would be helpful for future research to focus more on the mechanisms of effects and on which fathers and children benefitted most from FSSP. In this regard, we did not collect information on implementation of shared book reading at home; variability in implementation likely relates to outcomes following treatment. Integrating BPT and DR streamlined the intervention and created a context that may have aligned with fathers’ goals for their children, but the present study does not provide information about the relative contribution of these interventions. It is also unclear if siblings of children in the treatment group or other caregivers benefited from fathers’ participation. In addition, it is possible that fathers made gains in social support for their parenting through group participation, though this outcome was not investigated in the current study.

FUTURE DIRECTIONS AND CLINICAL IMPLICATIONS

The results of this trial are encouraging; however, questions remain regarding how to best engage and improve fathers’ parenting behavior and, ultimately, to support children’s functioning. First, it would important to ascertain the relative benefits of the key components of the FSSP program (i.e., DR, BPT). A study that includes a factorial design that allows for testing the key components of the FSSP program alone, as well as compared to other interventions (e.g., traditional BPT), will allow for determining what components impact which outcomes.

Relatedly, identifying moderators and mediators of the FSSP program is important. Although the study sample consists of families from low-income backgrounds, there is likely variability in key factors within these families (e.g., level of poverty; baseline level of problematic parenting; severity of child functioning) that may moderate response to the FSSP intervention. Moreover, the current trial did not collect quality data on engagement (within session and between sessions). Understanding the extent to which active engagement mediates outcomes appears to be an important area for further investigation (Chacko et al., 2016).

Future studies should also consider how best to involve and support mothers and fathers in improving their parenting and coparenting behavior. The development of the FSSP program was largely driven by our collaborative efforts with key stakeholders who recommended having father-only groups, which is in contrast to research suggesting that men may not be interested in this format (Russell et al., 1999). Our posttreatment focus groups suggest that many fathers appreciated the opportunity to connect with other fathers, whereas others noted that having the child’s mother more involved, in various ways, would have been preferred. This can be contrasted with a recent preference survey in which mothers from low-income families stated a preference for mother-only groups, whereas fathers were largely equivocal about group composition (Fabiano et al., 2016). Given that coparenting consistency and support is likely most beneficial for supporting children’s development (Teubert & Pinquart, 2010), consideration must be given to determine how and in what ways to involve mothers. Collectively, our experiences in developing and evaluating the FSSP program suggests that even with a relatively homogenous population of fathers, there are varying preferences for the parameters of the intervention. Although the core elements of the intervention are what mostly influences outcomes, the parameters (duration, intensity, format, etc.) are likely key factors in engaging populations in an intervention. Further empirical investigation into developing different “types” of interventions that vary in key parameters but maintain core elements of BPT may be necessary for different groups of parents.

Whereas the goal of this trial was not to determine the potential sustainability of the FSSP program, the development of the intervention had sustainability as a key goal. As we detailed earlier, deployment models (e.g., Weisz et al., 2005), where intervention development focused on “starting with the end in mind,” were a central framework for the development of the FSSP program. Our focus was to develop an effective, efficient, engaging, and sustainable model that could be feasibly implemented within the constraints of the practice setting. From our perspective, streamlining BPT through a focus on common elements of content and methods of delivery allowed for the development of a relatively simple and efficient intervention. Simple interventions that leverage the inherent strengths of group processes result in further lowering training and supervision demands. Focusing on outcomes (e.g., language skills to improve academic readiness) and contexts (e.g., shared book reading) that are relevant and meaningful for multiple stakeholders are likely critical for a full buy-in from the target population(s) and delivery setting. We believe that the field should continue to advance questions of intervention efficacy but also larger issues of effectiveness, sustainability, and scalability by considering key models (e.g., deployment focused, task shifting) and utilizing the science of common elements from the outset of treatment development to improve programs not only for mothers but also for fathers.

This study offers key suggestions for clinical practice. First, it is important that all individuals who interact with children in key contexts learn evidence-based approaches to supporting parents in order to support children’s social, emotional, and academic functioning. Anecdotally, our most successful facilitators were those with the least formal education; they were those individuals whom parents sought out most often for advice about their children. Understanding characteristics and skills that contribute to success as a facilitator is key for maximizing intervention impact. Preparing such individuals (in fact, all individuals) with the common elements of effective BPT practices may impact larger numbers of families over time through the informal interactions these individuals have with parents and children.

As discussed, programs that wish to engage parents should consider the key parent–child contexts that can serve as a context for learning parenting skills. As demonstrated in this study, shared book reading was an effective context in which to engage fathers in Head Start, but it was not widely accepted, given low rates of recruitment at some sites. Our experience in recruiting suggests that there are different segments within a population that may prefer different parent–child contexts as a vehicle for supporting parent–child interactions. Analytical methods, such as consumer preference methods (Fabiano et al., 2016), although often used in research, may allow practice settings to better identify key segments within their population that have distinct preferences. In this way, a few parent–child contexts that resonate with large segments of the population can be identified and used to integrate common elements of effective BPT practice. As we have recognized, one program will not engage an entire population, even when a population is relatively homogenous (i.e., fathers of children in Head Start in urban communities). Settings that recognize the diversity of preferences and leverage these preferences to engage their population will most likely be successful in broadly supporting parents.