A Forum for Researchers
Eastman's (1975) recent expression in the JRME of concern over so few replication studies is a concern that should be shared by all who have an interest in research. Replication is critical in assessing the “significance” of research results. A correlation coefficient of .20, which is statistically significant at the .01 level, will be much more “significant” if we can demonstrate by replication that the same result occurs again and again. By replicating, we help rule out the possibility that a Type I error (we rejected the null hypothesis when it was true) occurred in the original experiment. Furthermore, by independently replicating with different subjects, at different times and places, we also are helping to increase the generalizability of any “significant” results we do obtain.