| http://www.w3.org/ns/prov#value | - Several important confounding variables were not resolved in the studies, including baseline differences between completers and noncompleters, lack of blinding to treatment condition, a mixed approach to making condition assignments, and differential attrition across treatment groups.
|
