Enhancing Parenting Skills

P1: GCR Prevention Science [PREV]

pp575-prev-378528

August 24, 2002

12:51

Style file version Nov. 04, 2000

C 2002) Prevention Science, Vol. 3, No. 3, September 2002 (°

Enhancing the Parenting Skills of Head Start Families During the Transition to Kindergarten Elizabeth A. Stormshak,1,3 Ruth A. Kaminski,2 Matthew R. Goodman1

Head Start centers provide an excellent context for the implementation and success of familybased interventions, particularly home visiting. Based on a developmental–ecological model, a universal family-centered intervention was implemented with Head Start families. Outcome data from this parenting and home visiting program is presented (Project STAR: Steps to Achieving Resilience). Results suggest that both parenting groups and home visiting interventions are effective at enhancing parenting skills: however, home visiting programs have a higher participation rate. Additionally, home visiting by familiar staff was particularly successful at improving parenting skills at follow-up. Results suggest that embedding targeted interventions in universal strategies can be an effective means of engaging families in services. The results have implications for service delivery methods in early childhood as a means of enhancing parent participation. KEY WORDS: parenting intervention; home visiting; Head Start; early childhood.

Preparing children for school is one of the greatest challenges faced by educators today. As we enter the twenty-first century, the national educational goal that all children will enter school ready to learn is particularly challenging in light of the fact that more young children than ever before—approximately one in four—are living in poverty. Among families of young children there exists more homelessness, single-parent families, exposure to street violence, illegal drugs, and life-threatening illnesses such as AIDS. In addition, many families of young children lack affordable health services and child care. It is a consistent finding in the literature that children from low-socioeconomic backgrounds are significantly less likely than their middle class peers to enter school ready to learn and to achieve success (Entwisle &

Alexander, 1988; Hart & Risley, 1995; Landesman & Ramey, 1989). The developmental course of conduct problems in early childhood is complex. Based on an ecological model of development, children are embedded in a series of nested structures, each uniquely impacting their developmental course (Bronfenbrenner, 1979). Within this model, the family, school, and community all serve to either protect children from later problems or these systems contribute to the development of problem behavior by creating additional risk factors for children to overcome. Longitudinal, developmental research has identified a set of problem behaviors that place children at-risk in early childhood for later problems such as delinquency and substance abuse. These include aggressive behavior problems, social skill deficits, and achievement difficulties such as early reading problems. Early risk factors are exacerbated by contextual risks such as poverty. Moreover, these behaviors are associated with additional risk at transitional points in development (e.g., as children enter school). Ameliorating risks during critical junctures can alter the trajectories of children and prevent the subsequent development of problems (Coie et al., 1993). Interventions aimed at each level of the

1

Counseling Psychology Program, University of Oregon, Eugene, Oregon. 2 School Psychology Program, University of Oregon, Eugene, Oregon. 3 Correspondence should be directed to Elizabeth A. Stormshak, PhD, Counseling Psychology and Human Services, 5251 University of Oregon, Eugene, Oregon 97403-5501; e-mail: bstorm@ darkwing.uore.ed.

223 C 2002 Society for Prevention Research 1389-4986/02/0900-0223/1 °

P1: GCR Prevention Science [PREV]

pp575-prev-378528

August 24, 2002

12:51

Style file version Nov. 04, 2000

224

Stormshak, Kaminski, and Goodman

child’s social ecology are clearly the most effective at reducing problem behavior and enhancing social development.

PROJECT STAR (STEPS TO ACHIEVING RESILIENCE) Project STAR (Steps to Achieving Resilience: Kaminski & Stormshak, in press) was developed as a preventive intervention aimed at reducing the risk of substance use and delinquency in an at-risk sample of preschoolers (Head Start). The intervention targeted each level of the social ecology, and included a schoolbased universal intervention, a parenting group intervention, and home visits to families during the transition to kindergarten. Home visits were designed to enhance the content of the parenting group. Project STAR was developed based on the ecological model and was focused on reducing risk factors that lead to substance use in middle childhood. The targets of Project STAR included enhancing known risk factors such as social competence, self-regulation, and early literacy skills. These risk factors represented distal outcomes expected to emerge after targeting a more proximal outcome, which was parenting and family–school involvement.

Parenting as a Risk Factor There is probably no more important target in early childhood prevention than parenting. Defined broadly, “parenting” refers to skill-based practices as well as the quality of the parent–child relationship. In early childhood, relationship variables such as the quality of attachment, positive family involvement, and positive parent–child interactions predict both internalizing and externalizing behavior (Campbell, 1991). Positive parent–child relationships established in early childhood serve as a foundation for the utilization of effective parenting skills into the preschool years (Greenberg & Speltz, 1988). In general, parenting skills associated with later academic and behavioral problems include lack of caregiver involvement and positive parenting practices, poor and inconsistent family management, and punitive or negative parenting (e.g., Patterson & Stouthamer- Loeber, 1984; Petit et al., 1993; Stormshak et al., 2000). Patterns of interaction learned in the context of parent–child exchanges are typically generalized to school settings and peer groups, lead-

ing to the development of later problems such as drug use and delinquent behavior (Loeber et al., 1993). Behavior problems associated with this “early-starter” model and parenting skills contributing to these problems can be identified as early as age 2 and are predictive and stable into adolescence (Campbell & Ewing, 1990; Keenan & Shaw, 1994; Loeber et al., 1993). Research also indicates that parenting skills have an impact on school readiness and success. For example, in a longitudinal study of the cognitive performance, Burchinal et al. (1997) found that whether children’s cognitive performance increased or decreased during the preschool years was related to the responsiveness and stimulating characteristics of the child’s family environment. More recently, LambParker et al. (1999) found that parents understanding of play and facilitated learning predicted positive outcomes for Head Start children, whereas overcontrol and negative parenting predicted distractibility and hostility in the classroom and a subsequent decrease in early literacy skills. Similar findings were reported by Brody et al. (1994) in a study investigating the contributions of protective and risk factors to literacy and socioemotional competency in former Head Start children attending kindergarten. Brody and colleagues found that engaged, positive, and responsive caregiver–child interactions were strongly related to child social and academic achievement in kindergarten. Additionally, negativity in caregiver– child interactions was associated with lower socioemotional, literacy, and cognitive outcomes (Brody et al., 1994). Clearly, parenting interventions embedded in Head Start centers are a critical component of services for families to prevent mental health problems and promote school success.

Parenting in Context: Rural Oregon Children considered to be at especially high risk for behavior problems, deficits in social competence, and academic difficulties are often from economically disadvantaged homes (Hart & Risley, 1995; National Research Council, 1998). Project STAR was implemented in five different rural communities. These towns had been hard hit by unemployment and job instability, leading to poverty at a community level. Rural prevention research has been infrequent in the prevention literature, and often involves applying programs targeted at urban populations to rural populations with limited success. Recruitment and intervention can be particularly

P1: GCR Prevention Science [PREV]

pp575-prev-378528

August 24, 2002

12:51

Style file version Nov. 04, 2000

225

Enhancing Parenting Skills difficult when populations are dispersed geographically and the interventionists are seen as “outsiders” in the community (Spoth & Redmond, 1994). Research on the FAST Track Project (Conduct Problems Prevention Research Group [CPPRG], 1992) suggests that rural and urban differences in the implementation of prevention programs do exist and provides several guidelines for future prevention research with rural families. For example, rural parents are less likely to attend parenting groups and to talk freely in the context of other parents (Bierman & CPPRG, 1997).

Embedding Family Interventions in Existing Head Start Agencies Head Start is an agency that serves the many needs of at-risk children and families in early childhood. As such, it is an ideal setting for the delivery of family-centered interventions aimed at reducing future risk. Two family intervention components that have promise for implementation in Head Start agencies are parenting groups and home visiting. Both parenting groups and home visits have empirical support in the literature as effective interventions in early childhood (Olds et al., 1998; Webster-Stratton & Hammond, 1990). Parenting interventions, in particular, are associated with both changes in parenting skills as well as changes in child behavior (Serketich & Dumas, 1996). Both parenting groups and home visits are family interventions that occur within Head Starts currently; however, engaging parents in these activities is challenging. Webster-Stratton (1998) offered a parent training intervention to Head Start families in an urban setting. Attendance and participation were relatively good for this sample (88% of parents attended 50% of the parenting groups), and results suggested changes in parenting at posttest and follow-up. More recently, Webster-Stratton et al. (2001) found lower attendance rates for a briefer parenting group intervention in Head Start centers (51% of parents attended six or more sessions). A recent analysis of parent involvement in Head Start parent activities was conducted by Lamb-Parker et al. (2001). The researchers set out to understand the barriers to parent involvement in family-focused group activities in Head Start. They found that some of the biggest barriers to parental participation were maternal depression, having a baby or toddler at home, and having a schedule that conflicted with Head Start activities.

The results of this survey suggest that services other than group activities, such as home visiting, may be the best way to enhance family participation in Head Start activities.

Preliminary Research Over the course of our research on Project STAR, we have been funded by several grants to provide family services through Head Start. Family services were part of a comprehensive intervention that also included teacher training and classroombased intervention. The family-based services included three components: a parent training curriculum delivered in 12 and 6 sessions, home visits, and “family nights.” Family nights were individual Project STAR presentations addressing single topics that closely match the parent training and classroom curricula (e.g., playing with your child, establishing home routines). The Family Night presentations occurred at regularly scheduled Head Start parent meetings and, while not an intensive intervention, were an additional mechanism for engaging parents in Project STAR. Participation rates in Project STAR familybased interventions with Head Start agencies in rural Oregon across 6 years and three different grants have varied based on the service offered. Interestingly, the percentage of parents attending parenting groups was similar regardless of the number of sessions offered, with approximately one third of families selecting to attend the parenting groups. Home visiting had the highest percentage of participation (76%), followed by Family Nights (43%). In our initial research examining the efficacy of our intervention program, we compared outcomes of our intervention groups after the first and second year of our project. Analyses conducted after kindergarten compared parents who participated in our parenting program (parenting group and home visiting) to those who participated only in the universal intervention offered in classrooms. The results suggested that parents who participated in our parent group intervention with home visiting showed improvements in parenting skills (e.g., caregiver involvement) over the control group. Parents who participated only in our home visiting program had scores on caregiver involvement similar to those of our control group (Kaminski et al., in press). In other words, parent group attendance plus home visiting during the kindergarten year appeared to be the most effective treatment package for increasing parenting skills.

P1: GCR Prevention Science [PREV]

pp575-prev-378528

August 24, 2002

12:51

Style file version Nov. 04, 2000

226 As is true of most intervention projects, difficulties engaging families in treatment emerged in this study. Parenting groups for rural Head Start families were attended by only one third of the families even when child care, food, and transportation were provided and families were offered $5.00 for each parenting group attended. While almost three fourths of the families participated in the home visiting intervention, results of the home visiting intervention without parent group attendance were limited.

Stormshak, Kaminski, and Goodman we were interested in potential moderators and mediators of efficacy, including the relationship that parents had with their home visitor, the number of home visiting sessions, and risk factors such as public assistance, single parenting, and maternal depression.

METHODS Recruitment Recruitment of Schools

Transitioning children From Head Start to Elementary School: Building Relationships One major goal of Head Start is to show longterm benefits of this program into the high school years. Even with a comprehensive high quality Head Start program, advantages derived from preschool can be quickly lost without a smooth transition to a school environment and follow-up that builds upon previous gains (Lee & Loeb, 1995). One way to potentially enhance the short-term gains made by Head Start children would be to provide home visiting for the first year of elementary school. Ideally, these home visits would be conducted by someone who began a relationship with the families during the Head Start year. There is ample evidence within the community prevention literature as well as the psychotherapy research literature to suggest that clinical changes occur within the context of positive relationships. In community prevention research, collaboration with community leaders increases positive outcomes. Community problems are solved when the focus is on building programs that enhance existing strengths (see Weissberg & Greenberg, 1998, for a discussion of prevention and community research). In rural communities, existing strengths may include the relationship that families have with a known treatment provider such as Head Start staff. These relationships are an important component of continued services for families as children leave Head Start and enter elementary school. In Project STAR, parents were more likely to participate in our home visiting intervention than in our parenting group intervention. However, results of our home visiting intervention with no parent group attendance were limited. Given the success of our home visiting program in terms of family participation, we set out to more closely examine caregiver involvement as an outcome of our intervention. In particular,

Participating Head Start agencies in three rural counties of Oregon included 16 sites and 23 classrooms. Schools were recruited at the agency level and then randomly assigned by site to either intervention or control. Head Start sites with similar demographics were yoked and randomly assigned to different conditions. In many cases, Head Start sites contained multiple classrooms. Because teachers worked together in these sites, all classrooms within a site were assigned to the same condition. As a result, there were unequal numbers of classrooms assigned to the intervention and control group. After obtaining agency approval for the project, teachers were recruited to participate in the study. Recruitment was done at regularly scheduled Head Start staff meetings. Teachers did the primary recruitment of families during home visits or regularly scheduled meetings with parents. Project STAR staff supported teacher recruitment by attending parent meetings at regularly scheduled times and presenting the project. Two sites refused participation; hence, the final sample of schools included 14 Head Start sites with 11 classrooms assigned to the intervention and 9 classrooms assigned to the control condition.

Sample Four-year-old children and families from participating Head Start sites were recruited for this study. Of the 261 four-year- olds in the Head Start classrooms, 56% agreed to participate in Project STAR in Year 1 (n = 146). The final sample was 52% female, with 45% of children living in families including both biological parents. Sixty percent of the children and families received public assistance. Eighty-three percent were Caucasian and 13% were other ethnic groups, primarily Hispanic. These demographics are representative of the

P1: GCR Prevention Science [PREV]

pp575-prev-378528

August 24, 2002

12:51

Style file version Nov. 04, 2000

227

Enhancing Parenting Skills rural population in Oregon. Although we were unable to collect parent report data on the nonparticipants, 250 families consented to teacher reports of child behavior. There were no differences between children in the parenting portion of Project STAR and nonparticipating children on externalizing behavior problems or social competence in the classroom as reported by teachers on the Social Competence and Behavior Evaluation Scale (LaFreniere & Dumas, 1996); F(1, 250) = 2.15, p < .14.

in the classroom. During the fall of Year 1, teachers who agreed to participate attended a full-day workshop designed to train the teachers on administering this intervention, which consisted of circle time activities to promote critical skills in each of the four areas listed above. Head Start teachers then administered the intervention throughout the school year with consultation provided on a bimonthly basis from Project STAR staff.

Attrition

Selected Interventions

Data were collected over 2 years on our initial sample of 146 families (48 control families and 98 intervention). Between Wave 1 (fall of preschool) and Wave 2 (spring of preschool), we experienced 24% attrition, which resulted in a Year 2 sample of 112. These rates are similar to those reported by other researchers who have worked with Head Start populations (Webster-Stratton, 1998). When we examined the differences between participants who attrited and those retained at Wave 2, there were no differences in family make- up (maternal depression, marital problems), or child behavior problems. At Wave 3 (spring of kindergarten), our sample included 97 children and families, and hence, our attrition from Year 2 to 3 was only 13%.

Parents of participating 4-year-olds were offered a parenting group intervention in Year 1 (preschool) and a home visiting intervention in Year 2 (kindergarten). The parenting groups used the Incredible Years parenting curriculum developed by Carolyn Webster-Stratton for children aged 4–8 (Webster-Stratton, 1994). This video-based curriculum provides training in positive parenting, limit setting, problem solving, and handling misbehavior. The program has been implemented successfully in multiple research studies and leads to improved parenting as well as decreases in child behavior problems (Webster-Stratton, 1990; Webster-Stratton & Hammond, 1990). Parenting groups were run by trained interventionists at each participating Head Start site. Transportation and childcare were provided to families in order to increase participation. The home visiting intervention was developed as an individualized approach to working with high-risk families (Wasik et al., 1990). The home visiting curriculum combines the strategies from our classroombased intervention and the Incredible Years program into a home-based service delivery model (Kaminski & Stormshak, in press). The curriculum also includes other topics of relevance to children transitioning from preschool to elementary school (e.g., talking with your child’s teacher, asking your child about school, involvement in your child’s academic achievement). These additional components have been adapted from the FAST Track Parenting Intervention (CPPRG, 1992). At the initial meeting, families identified topics they were interested in from a list of choices. These included items such as early literacy and reading, social competence, parenting more effectively, and helping my child control anger. After this initial meeting, an individualized intervention package was developed with the interventionist and family. The overall goals of the home visiting were to provide

Self-Selection Random assignment occurred at the Head Start site level for this study, and hence, parents were self-selected into the various parenting interventions. However, there were very few preintervention group differences found on the variables of significance. Although parents who agreed to parenting groups had higher levels of school involvement, groups did not differ on caregiver involvement, behavior problems, or social competence (see Kaminski et al., in press). Intervention Procedures Universal School Intervention Children in participating classrooms received a classroom intervention administered by Head Start teachers. This intervention included 20 sessions focused on social competence, self-regulation, early literacy, and language and was administered to children by their Head Start teachers during regular activities

P1: GCR Prevention Science [PREV]

pp575-prev-378528

August 24, 2002

12:51

Style file version Nov. 04, 2000

228

Stormshak, Kaminski, and Goodman

support and education on significant predictors of behavior problems at this age. This included promoting social competence, self-regulation, parenting skills and family management, and academic success. The average number of home visits received by families was 5, with a range from 1 to 14. For purposes of analyses, families who received two or more home visits were considered to have participated in this aspect of the intervention.

of parent–child interactions were conducted at Waves 2 and 3 during the scheduled parent interview. Parents were paid $25.00 for each wave of assessment and $5.00 for each parenting group or home visiting session in which they participated.

Staffing

This questionnaire was developed based on items from the Discipline Questionnaire (Oregon Social Learning Center, 1992; LIFT Project) and the Parenting Practices Scale (Strayhorn & Weidman, 1988). The measure includes items assessing parenting warmth and praise, limit setting, problem solving, and punitive parenting strategies. In previous research with items from this measure, scales of warmth, consistency, punitiveness, and parental aggression (e.g., hitting) were formed in a confirmatory factor analysis and were differentially predictive of a variety of child behavior problems (Stormshak et al., 2000).

As a research project designed to be embedded in Head Start centers, Project STAR was focused on connecting with parents through Head Start. We took a community approach to our prevention program with a goal to train existing staff in our program content so that when our project was over, the staff would be able to use the materials and model in years to come. Limited resources in our rural communities made this community collaboration model even more critical. We hired and trained intervention staff alongside family consultants from the Head Start centers. Our staff worked collaboratively with these family consultants, teachers, and parents to coordinate our parenting groups, deliver our classroom-based intervention, and provide home visits to families. In each case, parenting groups were run with existing Head Start staff. We feel this collaboration enhanced overall participation in our parenting groups by creating a familiar environment for families and decreasing the likelihood that we would be seen as “outsiders” in these rural communities. Home visits were staffed by trained professionals, each with extensive experience working with families and children. Of the six home visitors, all had Master’s degrees in Counseling or Education. Each visitor participated in a full-day training workshop designed to improve the fidelity of the intervention. Home visitors met biweekly for supervision throughout the home visiting intervention. Assessment Procedure Parent interviews were conducted in the home during the three waves of data collection (fall preschool, spring preschool, and spring kindergarten). In each case, parents were interviewed by trained assessment staff and their responses were recorded on a laptop computer. Direct observations

Assessment Measures Parenting Practices Interview

Parent–Teacher Involvement Questionnaire (INVOLVE-P; CPPRG, 1999) This is a 26-item measure developed initially for the FAST Track program (CPPRG, 1992) that assessed the amount and type of contact that occurred between the parent and teacher in addition to the parent’s interest in talking with the teacher, satisfaction with the teacher, and involvement at school. This measure has good reliability and validity in addition to internal reliability of each subscale (school involvement and educational values, .89 and .91 respectively). Depression Scale (Radloff, 1977) This measure is based on the CES-D, and includes 20 items that provided a reliable and valid measure of self-reported depressive symptoms. This measure was included in the assessment battery to be used in analyses as a mediator. Parent-Child Interaction Parents and children participated in a 15-min interaction task modeled after the Behavioral

P1: GCR Prevention Science [PREV]

pp575-prev-378528

August 24, 2002

12:51

Style file version Nov. 04, 2000

229

Enhancing Parenting Skills Coding System developed by Forehand and McMahon (1981). The three 5-min tasks included child-directed play, parent-directed play, and a toy clean-up task. Trained observers used a modified version of the Coder Impressions Inventory (Capaldi & Patterson, 1989) and the Interaction Rating Scales (IRS; Crnic & Greenberg, 1990) to rate the quality of the parent–child interaction. Ratings included positive and negative parenting as well as child behavior.

MULTIRATER CONSTRUCT OF CAREGIVER INVOLVEMENT On the basis of previous research, we created a theoretically derived construct of caregiver involvement that consisted of two separate but related components, positive involvement and negative parenting. To test this theoretical model, a confirmatory factory analysis (CFA) was performed at each wave of data collection. In each CFA, six parent-report measures loaded onto the positive involvement construct, including the Involve-P Caregiver Involvement Scale, and five subscales of the Lift Parent Questionnaire (Positive Parent, Family Activities, Appropriate Limit Setting, Clear Expectations, and Monitoring). Direct observations of positive involvement also loaded onto the Positive Involvement construct. In each CFA there were two measures that loaded onto a Negative Parent construct: the Lift Harsh Discipline Scale and the Lift Inconsistent Parent Scale. Negative Parent was allowed to correlate with Positive Involvement in order to estimate the relation between these two constructs. The chi-square for this model was significant; however, the CFA models fit relatively well for all three waves as indicated by the Goodness of Fit Indices, (χ 2 (N = 146, d f = 16) = 25.79, p < .01, GFI = .96, TLI = .89. Additionally, all of the measurements significantly loaded onto the negative parenting or positive involvement constructs at all three waves, indicating that all the measures are good indicators of their respective construct. Finally, there was a moderate to strong negative correlation between the negative parenting and the positive involvement constructs (r = −.47, p < .01), suggesting that positive involvement and negative parenting are both subcomponents of caregiver involvement. Our final measure of caregiver involvement was one measure that included both positive and negative aspects of parenting.

RESULTS Results from our initial work with this sample seemed to indicate that the mechanism for change was the parenting group. That is, parents who attended the parenting group showed gains in caregiver involvement over those who attended the home visiting program only. However, despite positive results from the parenting group we had difficulty reaching the majority of our families with this service delivery model. Instead, families in our study appeared to prefer a home visiting approach to services as evidenced by our high rates of participation in this component. Our home visiting program was extensive and delivery was individualized across families. Because of random staff attrition, some families had home visits conducted by an unfamiliar staff person. In an attempt to understand potential moderators of efficacy associated with the home visiting program, we further subdivided our sample into those families who attended the parenting group and then received home visits from a familiar staff person (i.e., their parent group facilitator) versus those families who received home visits from a new intervention staff person (not the same person who facilitated their parent training group). These results are presented in Fig. 1. Although there were no differences between groups on Time 1 caregiver involvement scores, F(4, 91) = 1.03, MSE = 0.33, p = .39, all analyses controlled for Time 1 caregiver involvement. The results suggest that families who received the home visits from a familiar staff person had significantly higher caregiver involvement scores than every other group of families, F(1, 82) = 13.77, MSE = 2.88, p < 0.001. The results also suggest that families who received home visits from an unfamiliar staff person did not have significantly higher caregiver involvement scores than the other families, F(1, 82) = 0.39, MSE = 0.08, p = ns. Taken together, this analysis suggests that familiarity with the home visitor may be a critical ingredient in a successful home visiting program. The combined effect of a parenting group plus home visiting from a familiar staff person is the most effective set of interventions to increase caregiver involvement. Next we examined whether or not level of participation accounted for these effects. We looked at the number of home visiting sessions attended by each group of participants. Families who attended the parenting group prior to being offered the home visiting program participated in significantly more

P1: GCR Prevention Science [PREV]

pp575-prev-378528

August 24, 2002

12:51

Style file version Nov. 04, 2000

230

Stormshak, Kaminski, and Goodman

Fig. 1. Caregiver involvement by intervention group with familiar and unfamiliar project staff. Note. Class = Children who received only the classroom intervention; Class and Home Visit = Children and families who received the classroom and home visiting intervention with no parenting group. Unfamiliar versus familiar staff refers to the parents familiarity with their home visitor (familiar = same staffing as parenting-group intervention).

home visiting sessions than parents who did not attend the parenting group, F(1, 48) = 16.52, MSE = 88.13, p < .001. However, parents in the familiar versus unfamiliar group did not differ in their participation in home visiting. This finding suggests that level of participation in home visiting does not account for the influence of familiarity on caregiver involvement. However, a parenting group delivered prior to the home visiting intervention did increase family participation in the home visiting program. Next, we were interested in examining change in caregiver involvement as a function of number of home visiting sessions participated in across the three intervention groups (home visits only, parent group plus home visit from familiar staff person, and parent group plus home visit from unfamiliar staff person). We found a moderately significant interaction between intervention group and number of sessions attended on our caregiver involvement construct. This is shown in Fig. 2, F(3, 49) = 2.49, p = .083. Our limited sample size at this point decreased the power of our findings; however the trend depicted in Fig. 2 suggests that with a familiar home visitor, level of participation in home visiting, or dosage, predicted increases in caregiver involvement scores. It appears that for families with a familiar home visitor, more home visits led to increased improvements in caregiver involvement. However, with an unfamiliar visitor or no

parenting group, dosage did not predict caregiver involvement. For these families, dosage had no impact of levels of caregiver involvement. In an attempt to further understand potential moderators of efficacy associated with this program, we subdivided our sample into several final groups. We examined differences in caregiver involvement for families who received the home visiting intervention and were (a) on public assistance, (b) single parents, or (c) depressed. In separate analyses, we found no effects for public assistance or single parenting on our home visiting intervention. Parents who were on public assistance or who were single parents did not differ in their Wave 3 caregiver involvement scores from parents without these risk factors. When we subdivided our families by scores on the CES-D (using a mean split), we found that mother’s with more depression scored lower on Wave 3 caregiver involvement than mothers who were not depressed, F(1, 82) = 8.00, MSE = 2.53, p < .001. Both depressed and nondepressed parents received the same number of home visits (mean = 7). DISCUSSION This research presented findings from Project STAR, an early intervention program designed to

P1: GCR Prevention Science [PREV]

pp575-prev-378528

August 24, 2002

12:51

Style file version Nov. 04, 2000

231

Enhancing Parenting Skills

Fig. 2. Dosage of the home visiting program predicting caregiver involvement for familiar staffing.

embed family-centered services in an existing service delivery system for low income, rural families. It is clear from this research that home visiting interventions are an effective means of maintaining a connection with Head Start families as they transition from Head Start to kindergarten. Home visiting had the highest participation rate of all the interventions offered to Head Start families. Additionally, it appears that the staffing of this intervention is critical. Head Start families who were visited by a familiar staff person made significant gains above families who were visited by an unfamiliar staff person. These gains were not based on attendance. For families who were visited by a familiar staff person, there was a significant correlation between dosage (number of home visits) and caregiver involvement scores. This was not true of any other group. Although the literature suggests potential positive benefits to home visits in early childhood, the mechanism for change is still unclear. It may be, for example, that parents form a working relationship with the home visitor that allows for discussion and process around parenting issues as well as other family problems (Heinicke et al., 1999). In this case, the relationship with the home visitor plays a significant role in the change process. Support for this theory is also

found in research suggesting that parents who perceive the home visitor to be empathic will in turn show positive, empathic parenting with their infant (Korfmacher et al., 1998). It also may be that home visits allow the interventionist to understand cultural and contextual factors that may be impacting the child and family and provide a more sensitive intervention for that family (Slaughter-Defoe, 1993). Our findings are consistent with those of Olds et al. (1998, 1999) who suggested that home visits need to be of sufficient frequency and duration of visit need to be sufficiently long to effect change. In addition, as evidenced by our results, continuity of staff appears to be an important factor. This relationship-based enhancement of effects does not necessarily mean, however, that the relationship qualities of the home visitor were responsible for the changes in families. It may be that Head Start families take some time to build trust with staff and with their community. The continuity of a familiar staff person from Head Start to kindergarten provided a community connection for these families that was supportive and constant. In turn, families were able to form a working relationship more quickly with a familiar interventionist and perhaps work more intensely on parenting problems and family–school relationships.

P1: GCR Prevention Science [PREV]

pp575-prev-378528

August 24, 2002

12:51

Style file version Nov. 04, 2000

232 An added benefit of the home visiting model is that therapists have the opportunity to understand the family in their own context. Child development and behavioral concerns of parents can all be understood more fully in the situational contexts of the home visit (e.g., chaotic households, cramped living quarters, unsafe neighborhoods, and stressful interactions with neighbors and community). Although the majority of our sample was Caucasian, these nonspecific benefits of home visiting allowed staff to understand each family’s culture. Certainly, rural families do exist in their own culture. It is often times one of isolation, community problems, and lack of resources. In particular, limited recreational, education, and mental health services maintain problems for rural families (Sherman, 1992). Stressful economic conditions in rural areas lead to family problems, which in turn predict child behavior problems (Conger & Elder, 1994). These problems can be understood more fully when interventions occur in the home. Our research effort involved embedding our intervention into existing services that were already being offered at Head Start centers. Parenting groups and support are provided as a regular part of Head Start. Home visits are also provided, but vary widely in implementation. In order to increase participation in our parenting groups, we offered $5.00 in gift certificates for each group attended, childcare, and transportation. We coordinated these incentives with the Head Start centers involved in our intervention to increase the likelihood of success (for example, Head Start staff provided the childcare). We also reduced the number of parent group sessions from 12 to 6 over the course of our work with Head Start centers in order to increase parent participation. We hypothesized that offering less sessions would increase attendance by reducing the overall time commitment from families. Despite our efforts directed toward parent group attendance, participation in our parenting groups was still low. In our research, almost three fourths of the rural families participated in home visits compared to only one third who attended the parenting groups. Although rural communities may seem like the perfect environment for a parenting group, it has been our experience that the opposite is true. The fact that most of our families knew each other well (due to living in a small community) may have at times prevented them from attending a parenting group because they feared that their problems were not going to be confidential. Families are not anonymous in rural communities and parents have relationships with

Stormshak, Kaminski, and Goodman other families that extend back to their own childhood years and sometimes even further. A home visiting intervention, therefore, may meet the needs of rural families much better than of urban families who may be less connected to their communities. Clearly, familiar staff enhanced the results of our intervention. In hiring staff to work on our project, we first looked for skilled therapists from the rural community. In only one case were we able to hire someone in the community to serve as a parent group facilitator and home visitor. The reasons for this were varied, and primarily were related to a lack of available and qualified individuals in these communities. In smaller communities, qualified staff are already employed full-time and Head Start teachers and family consultants were not available to do additional home visits even when offered extra compensation. This was in part due to their teacher-union guidelines as well as a lack of outside time and staff commitment to Project STAR. Thus, our home visiting staff were typically not from the community; however, they were familiar because we tried to use the same staff to run parenting groups and conduct the home visits 1 year later. By the time the staff were conducting the home visits, they had been working in the community for a full year. Staff attrition was not atypical and included factors such as moving and career changes. These normative reasons for staff attrition, however, can clearly change the results of a prevention program. This is a critical consideration for staffing prevention research and administering prevention programs in rural communities. When we examined dosage as a predictor of outcomes, we found that dosage was only related to caregiver involvement for families who were visited by a familiar home visitor. Although the results should be interpreted with caution, this preliminary finding has implications for future service delivery. In traditional psychotherapy outcome research based on a medical model of service delivery, dosage is conceptualized as having a direct relation to outcomes. That is, the more therapy that is delivered the greater the outcomes for children and families. However, recent research suggests that within child and family treatment models, clinically significant change may be based on multiple factors including group dynamics and adherence to the program rather than quantity of sessions delivered (Dishion & Kavanagh, in press). For example, in a careful analysis of clinical cases treated at Oregon Social Learning Center, Weber (1998) did not find evidence for a dose–response relation between the number of sessions and changes in child

P1: GCR Prevention Science [PREV]

pp575-prev-378528

August 24, 2002

12:51

Style file version Nov. 04, 2000

233

Enhancing Parenting Skills behavior or parenting. Changes appeared to be a function of the interchange between the therapist and the client rather than a simple dose–response relationship (Stoolmiller et al., 1993). In a home visiting service delivery model, the relationship with the home visitor appears to be a critical component in the relation between dosage and positive outcomes for families. We also examined several risk factors as predictors of outcomes of our home visiting program. These included public assistance, single parenting, and depression. We did not find any differences in families on public assistance or single parents; however, we did find that parents who were more depressed scored lower on caregiver involvement after the intervention than those parents who were not depressed. This finding is consistent with other research in this area that suggests depression can mediate the impact of parenting interventions (Webster-Stratton & Hammond, 1990). It is likely that parents who are depressed require a much more extensive intervention program, aimed at their child as well as contextual factors in their lives that may be predicting depression (e.g., marital problems, substance use, etc). Our program did not target these problems specifically but focused more broadly on parenting and family–school linkages. In summary, understanding factors that moderate the efficacy of family interventions is critical to understanding how to improve implementation and engage families in treatment. It is important in prevention to establish relationships with existing agencies that serve at-risk populations. Head Start operates in some regards as a community agency, serving the multiple needs of many families. Retaining this family connection into the kindergarten year and supporting families as they transition to school may have positive benefits for both children and families. Although this research did not examine child outcomes, improvements in parenting are an important first step to making overall changes in child behavior. Continued research with this population will be important in order to solidify our understanding of effective family interventions at this age.

ACKNOWLEDGMENTS This research was funded by a grant from The Center for Substance Abuse Prevention (UR6SP07957) to the second author, a grant from the U.S. Department of Education, Office of Educational Research and Improvement (R305T990474), to the

second and first authors, and a grant from the U.S. Department of Education, Office of Special Education Programs (H324C990074), to the second and first authors. Thanks to all the staff at Lane County, LinnBenton, and Douglas County Head Start who assisted with this research. Special thanks to Berry Broadbent and Stacy Mendez for their dedication to this work and strong community connections with teachers and families. REFERENCES Bierman, K. L., & The Conduct Problems Prevention Research Group (1997). Implementing a comprehensive program for the prevention of conduct problems in rural communities: The FAST Track experience. American Journal of Community Psychology, 25, 493–514. Brody, G. H., Stoneman, Z., & McCoy, J. K. (1994). Contributions of protective and risk factors to literacy and socioemotional competency in former Head Start children attending kindergarten. Early Childhood Research Quarterly, 9, 407–425. Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and by design. Cambridge, UK: Harvard University Press. Burchinal, M. R., Campbell, F. A., Bryant, D., Wasik, B. H., & Ramey, C. T. (1997). Early intervention and mediating processes in cognitive performance of children of low-income African-American families. Child Development, 68, 935– 954. Campbell, S. B. (1991). Behavior problems in preschool children: Clinical and developmental issues. New York: Guilford Press. Campbell, S. B., & Ewing, L. J. (1990). Follow-up of hard to manage preschoolers: Adjustment at age 9 and predictors of continuing symptoms. Journal of Child Psychology and Psychiatry, 31, 871–889. Capaldi, D. M., & Patterson, G. R. (1989). Psychometric properties of fourteen latent constructs from the Oregon Youth Study. New York: Springer. Coie, J. D., Watt, N. F., West, S. G., Hawkins, J. D., Asarnow, J. R., Markman, H. J., Ramey, S. L., Shure, M. B., & Long, B. (1993). The science of prevention: A conceptual framework and some directions for a national research program. American Psychologist, 48, 1012–1022. Conduct Problems Prevention Research Group (1992). A developmental and clinical model for the prevention of conduct disorders: The FAST Track program. Development and Psychopathology, 4, 509–527. Conduct Problems Prevention Research Group (1999). Technical reports for the FAST Track assessment battery. Unpublished manuscript. Conger, R. D., & Elder, G. H. (1994). Families in troubled times: Adapting to change in rural America. Hillsdale, NJ: Aldine. Crnic, K. A., & Greenberg, M. T. (1990). Minor parenting stress with young children. Child Development, 61, 1628–1637. Dishion, T. J., & Kavanagh, K. (in press). Adolescent problem behavior: A family-centered intervention and assessment sourcebook. New York: Guilford Press. Entwisle, D. R., & Alexander, K. L. (1998). Facilitating the transition to first grade: The nature of transition and research on factors affecting it. The Elementary School Journal, 98, 351– 364. Forehand, R., & McMahon, R. J. (1981). Helping the noncompliant child: A clinician’s guide to parent training. New York: Guilford Press.

P1: GCR Prevention Science [PREV]

pp575-prev-378528

August 24, 2002

12:51

Style file version Nov. 04, 2000

234 Greenberg, M. T., & Speltz, M. L. (1988). Contributions of attachment theory to the understanding of conduct problems during the preschool years. In J. Belsky & T. Nezworski (Eds.), Clinical implications of attachment (pp. 177–218). Hillsdale, NJ: Erlbaum. Hart, B., & Risley, T. R. (1995). Meaningful differences in the everyday experience of young american children. Baltimore: Paul Brookes. Heinicke, C. M., Fineman, N. R., Ruth, G., Recchia, S. L., Guthrie, D., & Rodning, C. (1999). Relationship-based intervention with at-risk mothers: Outcomes in the first year of life. Infant Mental Health Journal, 20, 349–374. Kaminski, R., & Stormshak, E. A. (in press). Early intervention for the prevention of substance abuse in preschool children. In P. Tolan, J. Szcopoznik, & S. Sambrano (Eds.), Preventing substance abuse ages 3–14. Kaminski, R. A., Stormshak, E., Good, R. H., & Goodman, M. (in press). Outcomes from Project STAR: Preventing later substance abuse in preschool children. Journal of Addictive Behavior. Keenan, K., & Shaw, D. S. (1994). The development of aggression in toddlers: A study of low-income families. Journal of Abnormal Child Psychology, 22, 53–77. Korfmacher, J., Kitzman, H., & Olds, D. (1998). Intervention processes as predictors of outcomes in preventive home visitation programs. Journal of Community Psychology, 26, 49–64. LaFreniere, P. J., & Dumas, J. E., (1996). Social competence and behavior evaluation in children ages 3 to 6: The short form (SCBE-30). Psychological Assessment, 8, 369–377. Lamb-Parker, F., Boak, A. Y., Griffin, K. W., Ripple, C., & Peay, L. (1999). Parent–child relationship, home learning environment, and school readiness. School Psychology Review, 28, 413–425. Lamb-Parker, F., Piotrkowski, C. S., Baker, A. J., Kessler, S., Clark, B., & Peay, L. (2001). Understanding barriers to parent involvement in Head Start: A research-community partnership. Early Chilhood Research Quarterly, 16, 35–51. Landesman, S., & Ramey, C. T. (1989). Developmental psychology and mental retardation: Integrating scientific principles with treatment practices. American Psychologist, 44, 409–415. Lee, V. E., & Loeb, S. (1995). Where do Head Start attendees end up? One reason why preschool effects fade out. Educational Evaluation and Policy Analysis, 17, 62–82. Loeber, R., Wung, P., Keenan, K., Giroux, B., Stouthamer-Loeber, M., Van Kammen, W. B., & Maughan, B. (1993). Developmental pathways in disruptive child behavior. Development and Psychopathology, 5, 103–134. National Research Council (1998). Preventing reading difficulties in young children. Washington, DC: National Academy Press. Olds, D., Henderson, C. R., Cole, R., Echenrode, J., Kitzman, H., Luckey, D., Pettit, L., Sidora, K., Morris, P., & Powers, J. (1998). Long-term effects of nurse home visitation on children’s criminal and antisocial behavior: 15-year follow-up of a randomized controlled trial. JAMA, 280, 1238–1244. Olds, D. L., Henderson, C. R., Harriet, J. K., Eckenrode, J. J., Cole, R. E., & Tatelbaum, R. C. (1999). Prenatal and infancy home visitation by nurses: Recent findings. The Future of Children, 9, 44–65. Oregon Social Learning Center (1992). Discipline Questionniare. Technical report for OSLC. Unpublished manuscript.

Stormshak, Kaminski, and Goodman Patterson, G. R., & Stouthamer-Loeber, M. (1984). The correlation of family management practices and delinquency. Child Development, 55, 1299–1307. Petit, G. S., Bates, J. E., & Dodge, K. A. (1993). Family interaction patterns and children’s conduct problems at home and school: A longitudinal perspective. School Psychology Review, 22, 403–420. Radloff, L. (1977). The CES-D Scale: A self-report depression scale for research in the general population. Journal of Applied Psychological Measurement, 1, 385–401. Serketich, W. J., & Dumas, J. E. (1996). The effectiveness of behavioral parent training to modify antisocial behavior in children: A meta-analysis. Behavior Therapy, 27, 171–186. Sherman, A. (1992). Falling by the wayside: Children in rural America. Washington, DC: Children’s Defense Fund. Slaughter-Defoe, D. T. (1993). Home visiting with families in poverty: Introducing the concept of culture. The Future of Our Children, 3, 172–183. Spoth, R., & Redmond, C. (1994). Effective recruitment of parents into family-focused prevention research: A comparison of two strategies. Psychology and Health: An International Journal, 9, 353–370. Stoolmiller, M., Duncan, T., Bank, L., & Patterson, G. (1993). Some problems and solutions in the study of change: Significant patterns of client resistance. Journal of Consulting and Clinical Psychology, 61, 920–928. Stormshak, E. A., Bierman, K. L., McMahon, R. J., Lengua, L., & The Conduct Problems Prevention Research Group (2000). Parenting practices and child disruptive behavior problems in early elementary school. Journal of Clinical Child Psychology, 29, 17–29. Strayhorn, J. M., & Weidman, C. S. (1988). A parent practices scale and its relation to parent and child mental health. Journal of the American Academy of Child and Adolescent Psychiatry, 27, 613–618 Wasik, B. H., Bryant, D. M., & Lyons, C. M. (1990). Home-visiting: Procedures for helping families. Newbury Park, CA: Sage. Weber, F. D. (1998). The dose–effect relationship in family therapy for conduct disordered youth. Unpublished doctoral dissertation, University of Oregon. Webster-Stratton, C. (1990). Long-term follow-up of families with young conduct problem children: From preschool to grade school. Journal of Clinical Child Psychology, 19, 1344–1349. Webster-Stratton, C. (1994). Advancing videotape parent training: A comparison study. Journal of Consulting and Clinical Psychology, 62, 583–593. Webster-Stratton, C. (1998). Preventing conduct problems in Head Start children: Strengthening parenting competencies. Journal of Consulting and Clinical Psychology, 66, 715–730. Webster-Stratton, C., & Hammond, M. (1990). Predictors of treatment outcome in parent training for families with conduct problem children. Behavior Therapy, 21, 319–337. Webster-Stratton, C., Reid, J., & Hammond, M. (2001). Preventing conduct problems, promoting social competence: A parent and teacher training partnership in Head Start. Journal of Clinical Child Psychology, 30, 282–302. Weissberg, R. P., & Greenberg, M. T. (1998). Prevention science and collaborative community action research: Combining the best from both perspectives. Journal of Mental Health, 7, 479–492.