ABSTRACT
Focused studies have addressed specific clinical syndromes and specific treatment approaches. Such research probably leads to the most important question: What helps which child, in what kind of family situation, with which diagnosis? However, methodologic problems including small sample size, absence of controls, unreliable measures, and differences in clinically and research relevant measures, all strikingly highlight the basic problem inherent even with meta-analysis. In fact, many research designs are relatively poor at discriminating moderate advances (Cohen, 1977). There are serious concerns about combining data, mixing different kinds of subjects, and the design problems of the individual studies. It is here that the old dictum: "Poor Data In, Poor Findings Out" is as true as ever.
