JUNTOS Familia Preventive Intervention

SCHOOL SELECTION FOR QUASI-EXPERIMENTAL GROUP DESIGN. For aim 4, we have selected two out of five high schools that are closely are closely matched on the following criteria: 1) % Latino student enrollment (85.6 and 84.7), 2) % of students classified as economically disadvantaged (83 and 87), 3) Student achievement scores (66 and 69), 4) Student attendance scores (92 and 93), and 5) Student mobility rate (25.8 and 25.1). The matching of the two pilot study schools on these criteria, will allow us to eliminate any potential confounds that might preclude us from teasing apart the effects of the intervention from potential confounding school level effects. If the proposed intervention shows promise, we will be able to systematically test for school level effects in a larger, future study. We then propose to randomly assign one of the two schools to serve as intervention school and one of the two schools to serve as comparison school. The random assignment of one school to serve as intervention and one school to serve as comparison school will demonstrate the feasibility of randomization for a future larger study and will eliminate concerns about crossover and contamination.

Families from the intervention school (N = 30) will then be assigned to one of three intervention groups, each consisting of about 10 families (i.e., one youth and one parent), based on their language preference (English vs. Spanish). These families will then participate in the 9 week pilot intervention and fill-out a pre- and post-assessment battery.

Families from the comparison school (N = 30) will receive weekly information packages and fill out the pre- and post-assessment battery in the same week as the intervention school families.

For both groups, we will aim towards equal representation of boys and girls as well as U.S. and foreign-born youth. In line with the academic calendar of all the schools, we will aim to deliver the intervention from January to May to avoid delays due to school holidays and vacation.

Intervention Families will receive intervention sessions designed in PHASES 1-3. Sessions will be delivered in English or Spanish, based on participants' language preferences. We will conduct three groups (10 families and 2 facilitators/group). Although the time and location will be determined after study PHASE 3, based on my discussions with AISD, intervention sessions may be offered after school in a safe room. Incentives will include weekly non-monetary incentives, starting with small incentives (e.g., pens) and increasing the incentive value in later weeks (e.g., t-shirts). Each session will include a meal. Also, the time and location of the intervention will be determined based on the qualitative interview feedback. Participants will receive $30 for each completed assessment. Intervention session content will be delivered through short presentations, small/large group discussions, role-plays, exercises and practice of new skills. Each session will include home practice of new skills. Sessions will be digitally recorded. Recording will be used for future intervention refinement. Participants will be asked to respect the privacy of other group members. Facilitators: The PI will lead intervention sessions with a bilingual facilitator. The PI will, in consultation with co-primary mentors, Martinez and Marsiglia, train the facilitator by reviewing the intervention manual, emphasizing the purpose of each intervention component session and the need to not modify the manual. We will conduct mock intervention sessions and provide in-person feedback.

Comparison Families will receive 8 to 10 weekly information packages about the college application process. An information package may provide information about the application timeline and planning process. Participants will receive the information packages the same week the intervention families meet and sign a receipt for each package. Incentives: Comparison families will receive the same incentives as intervention families: weekly non-monetary incentives for picking up information packages and $30 for each completed assessment.

.2.

STATISTICAL DESIGN \& POWER CONSIDERATIONS.

We will evaluate the feasibility and acceptability of the intervention with the following measures: number of families screened and enrolled per week, retention rates of intervention and comparison families, duration of assessment surveys and missing data, proportion of families who complete protocols for intervention and comparison intervention families, attendance at intervention sessions, preferred language for the intervention and comparison families, intervention engagement and interest, and reports of perceived barriers and facilitators for intervention participation.

We will also estimate whether the intervention shows promise on improving youth and parent coping skills, family functioning, and parenting behaviors and whether it shows promise in preventing youth alcohol outcomes and cognitions. The primary outcome of interest is coping skills. We will estimate the effect of the intervention on coping skills separately for youth and parents. Secondary outcomes include family functioning, parenting behaviors, alcohol use/related risks and alcohol related cognitions. We will estimate the effect of the intervention on family functioning, parenting behaviors, alcohol use/related risk and alcohol related cognitions separately for youth and parents.

In all intervention impact analyses, student and family outcomes will be compared for the intervention and comparison groups using an intention to treat (ITT) approach. We will use a generalized linear modeling framework, where post-intervention outcome measures will be regressed on an indicator variable for intervention versus comparison group status, while statistically controlling for baseline levels of coping skills, ethnic discrimination, and demographic variables.

Intervention impact estimates will be translated into standardized effect size indices (e.g., standardized mean differences, risk ratios) to facilitate interpretation of the magnitude and direction of potential intervention effects. Multiplicative interaction terms will be added to the generalized linear models to explore possible differential effects of the intervention on coping skills, family functioning, parenting behaviors, and alcohol outcomes and cognitions based on dosage and demographic variables.

Missing data will be handled with multiple imputation or full-information maximum likelihood-based approaches, presuming the data can be reasonably assumed to be missing at random. Should missing data or attrition be associated with covariates, appropriate covariates will be included in all impact analyses.

Given the exploratory nature of the proposed study, the moderate sample size, and a desire to detect early signs of promise of the intervention, we will focus on effect size estimates (e.g., 95% confidence intervals, standardized mean differences, Cohen's f), rather than on statistical significance. Co-primary mentor Martinez was able to identify beneficial effects of Nuestras Familias on parenting and youth outcomes with a sample of 73 families and such we believe that our sample size of 60 families will allow us to detect signs of promise of the intervention. Focusing on effect size estimates in this pilot feasibility study will allow me to plan a formal test of the intervention in a randomized trial (R01) to assess the efficacy of the intervention, isolate school level effects on intervention outcomes, and examine mediating and moderating mechanisms.