“…Moreover, the GMA effect size is independent of the intercorrelations among the repeated
measures over time, which meta-analysts have shown to be an important criterion for the
d from classical analysis of repeated-measures data to be expressed in
the same metric as the d calculated from independent (completely
randomized) groups designs (Becker, 1988; Dunlap, Cortina, Vaslow, & Burke, 1996; Morris, 2008; Morris
& DeShon, 2002). The GMA d , as calculated with Equation 1, is now frequently found in studies of
intervention efficacy, particularly in JCCP (e.g., Aderka, Gillihan, McLean, & Foa, 2013; Arch, Eifert, Davies, Vilardaga, Rose, & Craske, 2012; Chaffin, Funderburk, Bard, Valle, & Gurwitch,
2011; Kerr, DeGarmo, Leve, & Chamberlain,
2014; Ljótsson et al, 2013; Safren, O'Cleirigh, Bullis, Otto, Stein, &
Pollack, 2012; Twohig, Hayes, Plumb, Pruitt,
Collins, Hazlett-Stevens, & Woidneck, 2010). …”