Several caveats to the current findings should be noted. First, the
overall sample of 22 RCTs of SST was small, and the number of
studies with data for each of the outcome domains was even
smaller. Thus, as the number of RCTs of skills training grows in
the future, the findings of this meta-analysis will require replication.
Second, the vast majority of studies included in this metaanalysis
failed to report crucial sample information, such as age of
illness onset (15 of 23) and dosage and/or type of medication (15
of 23), limiting the number of variables that could be investigated
in this analysis and suggesting that some variables that may have
had an impact on SST effects remain to be explored. Third, some
of the moderator variable analyses included a very small numbers
of studies (e.g., only three of the seven SST studies included
clinic-based measures of functional capacity reported sample duration
of illness), making the power to detect some relationships
limited. Fourth, as is common to all meta-analyses, it is unknown
the degree to which our findings may represent publication bias.
Inclusion of unpublished negative findings would affect our overall
results, and their absence may have led us to overestimate our
reported effect sizes.