Research synthesis consists of systematic reviewing and meta-analysis. What can and can't these approaches do to synthesize results to reach general conclusions about scientific questions? What can we expect, and where do our expectations fall short? Are there alternatives? I will discuss some of the philosophical and practical differences and similarities between different disciplines in how these powerful tools are used, where developments are headed, and how they can and can't address risk. (Spoiler alert: meta-analysis is probably a lot better at interpolation than at extrapolation, but it's been underutilized in parameter estimation).
Dr. Jessica Gurevitch is a Distinguished Professor in the Department of Ecology and Evolution at Stony Brook University, a research university of the State University of New York. Dr. Gurevitch’s research interests span several traditional categories within the field of ecology, including research synthesis and meta-analysis, biological invasions, and broadly in ecology with a focus on plant communities and populations. She introduced contemporary quantitative research synthesis and meta-analysis to the fields of ecology and evolution, changing the way scientists in these fields conceptualize and review scientific data. This work has been controversial and highly influential, and grew out of her interests in applying rigorous statistical methodology to the analysis of ecological data and the design of ecological experiments.
In addition to carrying out scientific studies, Prof. Gurevitch has co-authored and co-edited several books, including Design and Analysis of Ecological Experiments (Scheiner and Gurevitch 1993, Chapman and Hall; 2nd ed. 2001, Oxford University Press), The Ecology of Plants (Gurevitch, Scheiner and Fox, Sinauer Assoc. 2002, 2006, 3rd ed. 2020—just out) and Handbook of Meta-analysis in Ecology and Evolution (2013, Koricheva, Gurevitch and Mengersen, Princeton University Press). In addition, Prof. Gurevitch co-authored an early software package for meta-analysis in ecology ( MetaWin 2.0 , Rosenberg, Adams and Gurevitch, publ. Sinauer Assoc.) as well as an open-access package, OpenMEE (2013, with several collaborators).