Figure 5

Autocorrelations in respect to the time variations of different features for the PHES (a)–(c), CHES between 2006 and 2011 (d)–(f), and CHES between 2012 and 2017 (g)–(i) networks. Panels show the autocorrelations computed for the temporal variations of demand-supply ratio (a), (d), (g), application scores (b), (e), (h), and gender balance levels (c), (f), (i). We identify positive autocorrelations up to two degree of separation on demand-supply and application scores, the same is not shown in the gender balance. Bars in light colors indicate an autocorrelation that is not significantly different from zero (failed a t-test with \(p > 0.05\))