Skip to main content
Fig. 4 | BMC Medicine

Fig. 4

From: The long-term health effects of attending a selective school: a natural experiment

Fig. 4

Assessment of the effect of selective schooling on outcomes derived from hospital and death records. Population is the study sample (n = 5039) and points are means of 50 people. a, d, g The first column shows the 11+ exam score vs the outcome. b, e, h The second column shows the same data with the population separated by selective (open orange triangles) and non-selective (open blue circles) secondary school attendance. The line shows the estimated effect size of secondary school attendance at the cut-point extrapolated across the optimal bandwidth. c, f, i The third column shows re-estimates of the effect size of secondary school attendance for a range of population sizes around the cut-point (bandwidth sizes). The effect size estimate at the predicted optimal bandwidth (reported in Table 4) is shown as a dashed line

Back to article page