B. T. Johnson. B, Mullen, and E. Salas (1995) concluded that the J. E. Hunter and F. L. Schmidt (1990) meta-analysis methods yield anomalous results in a variety of analyses. An examination of their procedures reveals that most of their conclusions to this effect stem from their use of a formula for the standard error of the mean correlation that is not used in the Hunter and Schmidt (1990) methods. Results produced using the appropriate formula are not anomalous and are similar to those from other meta-analysis procedures.
机构:
Mem Sloan Kettering Canc Ctr, Dept Epidemiol & Biostat, New York, NY 10021 USAMem Sloan Kettering Canc Ctr, Dept Epidemiol & Biostat, New York, NY 10021 USA