Fig. 2From: The long-term health effects of attending a selective school: a natural experimentProbability of attending a selective secondary school. Population is the study sample (n = 5039) with points showing mean scores for groups of 50 people. Dashed lines show the primary and secondary cut-points of 540 and 560 (rescaled to 0 and 20). Red lines are predicted probabilities of attending a selective secondary school within the population bandwidth of 76 points to either side of the primary score cut-pointBack to article page