Some studies reported more than one finding for a given outcome area. Some of
these multiple findings resulted from the use of more than one experimental or
control group in a single study. Others resulted from the use of several subscales
and subgroups in measuring a single outcome. Using several different ESs to
represent results from one outcome area in one study seems inappropriate to us.
The multiple ESs are not independent; they often come from a single group of
subjects or overlapping subject groups, and in any case they represent the effects
of a single program implemented in a single setting. To represent a single outcome
by several ESs violates the assumption of independence necessary for many
statistical tests and also gives undue weight to studies with multiple groups and
multiple scales.