Dr. Johnson, professor of psychology at the University of Connecticut, studies factors in the psychology of health promotion, and the theory and practice of research synthesis with a special emphasis on incorporating multilevel structures with meta-analysis. Ongoing projects include an NIMH-supported project, Syntheses of HIV & AIDS Research Project (SHARP), which is a series of meta-analyses evaluating the efficacy of HIV prevention interventions as a function of both intervention features and aspects known about the geographical area in which the study was conducted; and methods to reduce depression ranging from medicine to exercise.
This methodological presentation on meta-analysis will highlight best practices in model choices and depicting results, and offer practical advice and additional resources (e.g., software and websites). In many sciences, meta-analysis is routinely used to summarize trends in independent studies, using features of the study methods and their samples in an effort to explain differences in studies. Meta-analytic techniques have seldom brought community-level variables (e.g., cultural and economic features) to bear on study results even though, in theory, they are often relevant, and in practice, there are increasingly nuanced ways to do so. This methodological talk will emphasize new strategies for incorporating features of multi-level structures in the meta-analysis and highlight the potential role that stigma plays in relation to the efficacy of HIV prevention trials for African Americans.
Blair T. Johnson
Professor of Psychology
University of Connecticut