ON THE LOGIC AND PURPOSE OF SIGNIFICANCE TESTING

There has been much recent attention given to the problems involved wi th the traditional approach to null hypothesis significance testing (N HST). Many have suggested that, perhaps, NHST should be abandoned alto gether in favor of other bases for conclusions such as confidence inte rvals and effect size estimates (e.g., Schmidt, 1996). The purposes of this article are to (a) review the function that data analysis is sup posed to serve in the social sciences, Cb) examine the ways in which t hese functions are performed by NHST, (c) examine the case against NHS T, and (d) evaluate interval-based estimation as an alternative to NHS T.