BACKGROUND
Dropping out of high school is associated with numerous detrimental consequences, including low wages, unemployment, incarceration, and poverty. There are a large number of school and community-based prevention and intervention programs for general population and at-risk students, and there are a number of programs designed specifically to encourage school completion among pregnant and parenting teens. No comprehensive systematic reviews have examined these programs’ overall effectiveness.
OBJECTIVES
The objectives of this systematic review were to summarize the available evidence on the effects of prevention and intervention programs aimed at primary and secondary students for increasing school completion or reducing school dropout. The primary focus of the meta-analysis was to examine the comparative effectiveness of different programs and program approaches in an effort to identify those with the largest and most reliable effects on school completion and dropout outcomes. We also sought to summarize the effects of programs designed for pregnant and parenting teens.
SEARCH STRATEGY
A comprehensive and diverse international search strategy was used to locate qualifying studies reported between 1985 and 2010. A wide range of electronic bibliographic databases were searched, along with research registers, other grey literature databases, reference lists of all previous meta-analyses and reviews on the topic, as well as citations in research reports. We also maintained correspondence with researchers in the field of school dropout prevention.
SELECTION CRITERIA
Studies eligible for inclusion in the review were required to meet several eligibility criteria. First, studies must have involved a school-based or affiliated psychological, educational, or behavioral prevention or intervention program intended to have beneficial effects on students recipients; or, evaluated a community-based program explicitly presented as a dropout prevention or intervention program. Second, studies must have investigated outcomes for interventions directed toward school-aged youth, or those expected to attend pre-k to 12th grade primary and secondary schools or the equivalent. Third, studies must have used experimental or quasi-experimental research designs, including random assignment, non-random assignment with matching, or non-random assignment with statistical controls or sufficient information to permit calculation of pre-treatment effect size group equivalence. Fourth, studies must have reported at least one eligible outcome variable measuring school completion, school dropout, high school graduation, or school attendance. Finally, the date of publication or reporting of the study must have been in 1985 or later.
DATA COLLECTION AND ANALYSIS
The literature search yielded a total of 23,677 reports, 2,794 which were deemed potentially relevant and retrieved for eligibility determination. Of those, 548 reports describing 167 different studies were included in the final review. Random-effects inverse variance weighted meta-analytic methods were used to synthesize odds ratios for the school dropout outcomes. Meta-regression models were used to examine the effects of program characteristics, methodological characteristics, and participant characteristics on the dropout odds ratios. Funnel plots, regression tests for small sample bias, and trim and fill analyses were used to assess the possibility of publication bias.
RESULTS
General dropout programs (152 studies; 317 independent samples) and dropout programs for teen parents (15 studies; 51 independent samples) were analyzed in separate meta-analyses. Overall, both general dropout programs and programs specialized for teen parents were effective in reducing school dropout (or increasing school completion). The random effects weighted mean odds ratio for the general programs was 1.72. Using the average dropout rate for control groups of 21.1%, the odds ratio for the general programs translates to a dropout rate of 13%. For the teen parent programs, the mean odds ratio for gradu tion and dropout outcomes was 1.83 and was 1.55 for school enrollment outcomes. The average graduation rate for the young women in comparison groups was 26%. The corresponding graduation rate for young mothers in the intervention programs was 39%. For school enrollment outcomes, the average enrollment rate for the comparison mothers was 33%. The mean odds ratio of 1.55 for these studies translates into an enrollment rate of about 43%. Moderator analyses for the general programs indicated that studies with similar program and comparison groups at baseline and those that provided posttest data adjusted for baseline non-equivalence produced smaller effect sizes. For teen parent programs, moderator analyses found that random and matched designs produced smaller effect sizes than non-random or non-matched designs. Effect sizes were therefore adjusted for methodological characteristics to examine the effects of different program types net of the influence of method. The effects were generally consistent across different types of programs and for different types of participant samples. However, higher levels of implementation quality tended to be associated with larger effects. Analyses provided no strong indication of the presence of publication or small study bias.
AUTHOR’S CONCLUSIONS
Overall, results indicated that most school- and community-based programs were effective in decreasing school dropout. Given the minimal variation in effects across program types, the main conclusion from this review is that dropout prevention and intervention programs, regardless of type, will likely be effective if they are implemented well and are appropriate for the local environment. We recommend that policy makers and practitioners choosing dropout prevention programs consider the cost-effectiveness of programs, and choose those that fit best with local needs as well as implementer abilities and resources.
Oversett med Google Translate
