Fig 1.
Plotted coefficients for fixed effect covariates from a multilevel-multivariate analysis of drivers of school attendance for Tanzania.
Odds ratios and 95% confidence intervals from the two-level logistic regression model of school attendance among children of adolescents or youth-level using DHS data in Tanzania (2015–16 DHS, N = 6,197). Fixed effects only were plotted. A map of the regions of Tanzania is presented in S2 Fig.
Fig 2.
Geographic variation in theoretical travel time to the nearest secondary school.
The blue dots are locations of DHS clusters (n = 595); Light blue areas are large water bodies extracted from the open source repository DivaGis (www.diva-gis.org). DHS clusters, also called primary sampling units (PSUs) are typically census enumeration areas (EAs) which form the survey cluster.
Fig 3.
Shows the mean probability to access secondary school for each category of household’s wealth quintile, by (a) travel time to the nearest secondary school; b) level of education of the household’s head; c) child’s sex; d) shows the mean probability to access secondary school for any additional child under the age of five, by level of education of the household’s head.