Figure 13 : Gap in application rates between learners from Uni Connect areas and the average application rate of 1,000 matched counterfactual groups of learners from non-Uni Connect areas
113 . However , Figure 13 above does not show the statistical uncertainty around these estimates . It only shows the gap in application rates between learners from Uni Connect areas and the average of all 1,000 matched counterfactual groups . For some of the matched counterfactual groups , the change in the gap was larger , while for others it was smaller .
114 . Figure 14 below , known as a ‘ violin plot ’, shows the full range of estimated changes in the application gap across all 1,000 matched counterfactual groups within each cohort . The shaded areas show the range of estimated changes in the gap for each year since 2016 , with wider sections indicating a higher probability that the change in the gap is equal to the value along the vertical axis . The labels in bold represent the average estimated change in the gap , which correspond exactly to the change in the gaps since 2016 , which are shown in Figure 13 above .
115 . To give a sense of the range within which we can be confident that the true value lies , the upper and lower labels on the chart indicate the 25 th and 975 th estimates in order of size . These
41