Pre-post effect sizes should be avoided in meta-analyses
P. Cuijpers, E. Weitz, I. A. Cristea, J. Twisk
Epidemiology and Psychiatric Sciences · 2017
10.1017/S2045796016000809