Project Activities
The meta-analysis focused on students in kindergarten through high school (K–12) and considered school-based marginalization processes, such as interpersonal discrimination, disparate disciplinary actions or policies, students' perceptions of fairness of discipline, and peer victimization based on social position characteristics. The research team identified relevant articles, developed a system for coding, and used a two-stage approach to meta-analytic structural equation modeling to pool together the relevant correlation matrix elements and sample characteristics that were then used to test the research questions.
Structured Abstract
Setting
Studies included in the meta-analysis had to include students in grades K–12 enrolled in a public or private elementary or secondary school. Studies of home-schooled children, those in early childhood education programs, and college students were excluded. Studies from any country meeting these criteria were eligible for inclusion if results were provided in English.
Sample
The meta-analysis focused on studies of K-12 students. Students, educators, and school administrators were the reporters of school-based marginalization.
Factors
Malleable factors included marginalization within schools, student social relationships with peers and teachers, and student social behavioral competencies.
Research design and methods
To identify all possible studies on school-based marginalization, achievement, and social behavioral competencies, the research team searched relevant literature using four complementary strategies: reference database searches, antecedent searches, descendent searches, and grey literature searches. Studies included in the meta-analysis must have presented results pertaining to school-based marginalization, including discriminatory, racist, or prejudicial treatment (tied to various social positions, including race or ethnicity, socioeconomic status, nativity, LGBTQ status, gender); disparate, inconsistent, or harsh disciplinary strategies; and general unfair treatment or unfair rule implementation. The research team focused on social and psychological forms of marginalization that are more frequently tied to students' social position characteristics and are also likely to include school personnel as perpetrators. The studies also must have included statistics (descriptive or inferential) that could provide an estimate of the correlation between included focal constructs. The studies' research designs most commonly were correlational studies, but studies with experimental or quasi-experimental designs were eligible for inclusion. Designs could have been cross-sectional, where all measures are collected at a single time point, or longitudinal, where measures are assessed repeatedly across time (longitudinal associations between central constructs of interest may be within or across time).
Control condition
Due to the exploratory nature of this study, there was no control condition.
Key measures
School-based marginalization included teacher- and peer-perpetrated discrimination, disparate discipline actions and policies, students' perceptions of unfair or harsh treatment, and peer victimization based on social identities. Social behavioral competencies most typically included school engagement (behavioral, cognitive, emotional), on/off-task behaviors, academic motivation, effort/persistence, value of education, educational expectations or aspirations, self-related perceptions (efficacy, esteem, worth, concept), truancy, and discipline referrals. Achievement outcomes most commonly included grades or grade point averages, teacher-designed tests, standardized achievement tests, and general ratings of academic performance. Focal mediators included student-teacher and student-peer relationships and social- emotional well-being. Target social position indicators included race/ethnicity, SES, gender, nativity, and LGBTQ status.
Data analytic strategy
The research team used a two-stage approach to meta-analytic structural equation modeling, controlling for experiment-wise type I error rates using Bonferroni's modification of the a-level used for testing each path in each structural equation model. They first synthesized studies' estimates of correlations between each pair of constructs. After synthesizing the set of correlations, the resulting pooled correlation matrix and its associated covariance matrix then provided the data that the research team analyzed using structural equation modeling techniques to address each meta-analytic research question.
Key outcomes
The main findings of this exploratory study will be shared once they are available in a publicly available peer-reviewed publication.
People and institutions involved
IES program contact(s)
Products and publications
ERIC Citations: Find available citations in ERIC for this award here.
Supplemental information
Co-Principal Investigators: Benner, Aprile; Beretvas, Natasha
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.