Social Capital and Children's Development: A randomized controlled trial conducted in 52 schools in Phoenix and San Antonio, 2008-2015 (ICPSR 35481)

Time Period(s) View help for Time Period(s)

Date of Collection View help for Date of Collection

Data Collection Notes View help for Data Collection Notes

Study Purpose View help for Study Purpose

The goal of this study was to measure the causal effects of social capital on levels and inequalities of children's social and cognitive development during the early elementary years.

Sample View help for Sample

This cluster-randomized controlled trial assigned 52 elementary schools in San Antonio, TX, and Phoenix, AZ, to FAST (n = 26) and no FAST (n = 26). Participants were 3,084 parents, 52.1 percent (n = 1,607) from San Antonio and 47.9 percent (n = 1,477) from Phoenix, whose first grade child was enrolled in a study school.

Cities and school districts were selected because they had effective social service agencies with histories of implementing FAST, the school district population included high proportions of low income Latino families, and they agreed to participate in the study (districts agreed to allow schools to be randomly assigned to treatment status).

Schools were selected by the school districts. Districts were asked to invite the principals of Title 1 schools with at least 25 percent Latino students (during the 2007/2008 school year) to attend a meeting where they were informed about the study and invited to consent to participate. Principals who attended the meeting consented to participate, but not every principal who was invited to the meeting actually attended. In a few cases, insufficient principals attended or the schools did not meet the criteria for inclusion, and the district was asked to recruit additional schools.

Students and families were selected based on enrollment in first grade at a study school during the first year of the study for that school (2008-2009 for Cohort 1; 2009-2010 for Cohort 2). All first grade students/families were eligible so long as they consented to participate in the study. Seventy-six percent of the sample was Latino, with other minority groups represented at lower levels: 8 percent African American, 1.6 percent Asian/Pacific Islander, and 1.3 percent American Indian. There were no statistically significant differences between recruitment rates across the FAST and comparison schools. Over 68 percent (N = 2,041) of parents completed post-test surveys. In Phoenix, 50 percent were born outside of the United States and 57 percent spoke Spanish as their native language. In San Antonio, 24 percent of parents were born outside of the United States and 31 percent spoke Spanish as their native language.

Time Method View help for Time Method

Universe View help for Universe

First grade students and their parents in 52 elementary schools in Phoenix, AZ and San Antonio, TX.

Unit(s) of Observation View help for Unit(s) of Observation

Data Type(s) View help for Data Type(s)

Mode of Data Collection View help for Mode of Data Collection

Response Rates View help for Response Rates

Just below 60 percent of first-grade families consented to participate in the study.

Presence of Common Scales View help for Presence of Common Scales

The study measured three aspects of social capital: intergenerational closure, shared expectations, and relational trust. The various strands of social capital are measured at four time points in parent surveys. These survey items, and other items measuring parent-teacher relationships and parent-school social capital, are adapted from three sources: FAST National pre-and post parent surveys, Shumow, Vandel and Kang (1996) and Bryk and Schneider (2002). Measures of children's socioemotional behavior come from parent reports and teacher reports through a series of questions from the Strengths and Difficulties Questionnaire (SDQ) (Goodman 1997). This widely used instrument for assessing social adjustment and behavior problems taps five dimensions of psychological functioning: emotional symptoms, behavior problems, hyperactivity/inattention, peer relationship problems, and prosocial behavior.

Parents were asked to provide information about acculturation by responding to the Abbreviated Multidimensional Acculturation Scale (AMAS; Zea, Asner-Self, Birman, and Buki, 2003). This measure assumes that acculturation is a process by which individuals retain characteristics of the culture of origin while simultaneously acquiring characteristics of the new culture. It consists of three subscales: cultural competence, language competence, and ethnic identity. The AMAS was developed in English and Spanish and has appropriate psychometric properties with both United States-born and immigrant populations (Zea et al., 2003).

In addition to the AMAS, the study assessed language preference by tracking parents' requests to complete the pre-test survey in English or Spanish. Participants were asked to report their length of residence in the United States on a scale of 0-5 years, 6-10 years, and 11 or more years. Length of residence is a common marker of acculturation (Harris, Firestone, and Vega, 2005; Schwartz, Pantin, Sullivan, Prado, and Szapocznik, 2006).

Parental depression was measured using a reliable brief depression measure designed to reduce burden to participants during data collection, the Patient Health Questionnaire-2 (PHQ-2) to assess parents' depressive symptoms because of its high construct and criterion validity with other depression measures and with measures of functional status (Kroenk, Spitzer, and Williams, 2003; Mitchell and Coyne, 2007). The PHQ-2 is comprised of two items that assess the frequency of anhedonia and depressed mood in the past 2 weeks, both commonly considered as the cardinal symptoms of depression.

Finally, children's academic outcomes were measured using third grade state exam test scores. Raw test scores and scale scores were collected from the school districts in San Antonio and Phoenix. Additionally the study included the scores converted to percentiles, z-scores, normal curve equivalents, and a score standardized by control group (i.e., participants in the comparison treatment condition) means by language and site.

Original Release Date View help for Original Release Date

2015-02-03

Version History View help for Version History

Weight View help for Weight

Cross-sectional non-response weights for the 1st grade parent post-test surveys and 3rd grade parent post-test surveys and panel weights for researchers wanting to investigate longitudinal questions. Nonresponse weights weight individual cases by the inverse of the "response propensity." The weights were calculated by predicting nonresponse using multilevel logistic regressions. The study balanced having a powerful predictive model with the introduction of extreme weights, which resulted from small sample sizes in particular categories. Therefore, iterated models were utilized in order to include only characteristics that significantly predicted nonresponse or improved model fit. Students who are missing item-level data on predictors included in the nonresponse model do not have nonresponse weights. Calculating nonresponse weights for these students would require imputing missing data, and the study did not impute these data because they likely were not missing at random.