Student- and Classroom-Level Predictors of Refugee Student Attendance Trajectories in Lebanon: Implications for Interventions

• In Event: 3-095 - Variation in implementation of social-emotional programs for refugee children in low-, middle-, and high-income countries

Integrative Statement

Since the beginning of the Syria conflict in 2011, Lebanon has faced a growing educational crisis as the public school system struggles to meet the needs of refugee children (Human Rights Watch, 2016). As of 2018, approximately 58% of Syrian children ages 3-18 in Lebanon were out of school (UNHCR, 2018). Reports further indicate that refugee students enrolled in Lebanese public schools are high risk for low attendance and dropout (Jalbout, 2015). Reports outline a number of barriers to refugee student attendance, such as financial strains, high mobility, transportation access, and negative social experiences at school (e.g., Dryden-Peterson, 2015). Inconsistent school attendance at both the student and classroom level may have important implications not only for individual students’ received dosage, but also for the effectiveness of interventions implemented in school settings. Universal interventions targeting student outcomes often also depend on group-level impacts that reinforce individual-level effects over time (Cook et al., 2014). Inconsistent school attendance could interfere with both individual- and group-level intervention effects. Risk factors for low school attendance among refugee students may thus have broader important implications for the implementation of interventions in these contexts. The present study therefore aims to identify student- and classroom-level risk factors predicting refugee student attendance over the course of the school year in Lebanon.
Data were collected as part of a large-scale randomized control trial in Lebanon public schools, which evaluated different social-emotional and educational support programs for Syrian refugee students. Participants were recruited at 21 different research sites located in the Akkar and Bekaa regions of Lebanon. Present analyses include data from the first year of data collection for this study, during which student attendance to a classroom-based social-emotional learning program was reported for nine months. Student and classroom characteristics and potential risk factors were assessed using administrative records and parent questionnaires. Parents surveys included questions regarding their children’s health, their perceptions of community safety, whether their family had moved in the past year, and the different technologies they had available in their households (e.g., phone, refrigerator).
The present study will use multilevel growth models to predict change in students’ monthly attendance rate across nine months. Preliminary analyses predicted student attendance during the first data collection cycle (a period of five months), using a subset of 2,991 students in 141 classrooms who joined the study by the first month of this cycle. Table 1 lists sample characteristics and descriptive statistics calculated for this subset. The preliminary, main effects model predicted monthly attendance from student and classroom characteristics, with random intercepts and slopes at both student and classroom level (see Table 2). Monthly attendance rates tended to decline over time. Attendance was initially higher for female students, those with scoring at higher baseline levels of Arabic, and higher levels of parent-reported community safety. Grade level, child labor, household technology, and living in the Bekaa region predicted lower attendance. Further analyses will test whether student and classroom characteristics differentially predict changes in student attendance trajectories over time.

Authors

• Abigail C. Keim, The Pennsylvania State University
  Presenting Author

• Ha Yeon Kim, Global TIES for Children, New York University
  Non-Presenting Author