WITHDRAWN: High School Dropout Prevention Programs: Differential Effects By Gender, Race/Ethnicity, and Socioeconomic Status

Introduction: Despite substantial declines over the past decades, high school dropout remains an enduring educational and social problem in the United States. According to the National Center for Education Statistics, in 2016, approximately 6.1 percent—roughly 2.3 million—of US youth aged 16-24 years were high school dropouts. Dropout rates tend to be higher among males, racial/ethnic minorities, and socioeconomically disadvantaged groups. Although research on dropout prevention programs abounds, most studies were conducted using small samples from a single school, district, or city, limiting their generalizability. This study examines the effects of attending high schools with dropout prevention programs on dropout status with a nationally representative sample of high school students.

Methods: The data source for this study was the US High School Longitudinal Study of 2009-2016 (HSLS:09/16). Participants were 12,046 first-time 9^th graders from 872 high schools, who participated in the base-year (2009) and second follow-up (2016) surveys. Independent variable was whether the high school the student attended had dropout prevention programs. Control variables were individual backgrounds (including gender, race/ethnicity, socioeconomic status [SES], immigration status, cognitive scores, and pre-high school academic preparation) and school characteristics (including sector, type, urbanicity, region, academic performance, and poverty level). Dependent variable was dropout status (by year 2016). Linear probability regression models that accounted for HSLS:09/16’s sampling and panel weights were employed. Robust standard errors were calculated by clustering at the school level to address data-dependency issues.

Results: While no overall main effect of attending high schools with dropout prevention programs on dropout was found, results from two-way interaction tests indicated that attending schools with dropout prevention programs decreased the probability of dropout for Blacks by 8.4 percentage points (p<.05) and for students from the bottom one-fifth of SES by 4.1 percentage points (p<.05). Three-way interaction results showed that enrolling in a high school with dropout prevention programs decreased the probability of dropout for male Blacks by 13.7 percentage points (p<.05) and for male students from the bottom one-fifth of SES by 15.0 percentage points (p<.05).

Conclusions: Evidence from this national study suggested that dropout prevention programs in US high schools may be effective in reducing dropout rates among students who were Black and from lowest-SES households, particularly males, but not among students who were Hispanic and from low-middle-SES homes. Future prevention polices/programs and research should address the needs of Hispanic and low-middle-SES students, who are also at relatively high risk of dropping out.