Defending the statistical status quo

Journal/Book: Theor Psychol. 1999; 9: 2455 Teller Rd, Thousand Oaks, CA 91320, USA. Sage Publications Inc. 183-189.

Abstract: The practices suggested by Thompson (1999) have not been adopted by researchers because they would rarely be useful. There are problems with reports of effect size that depend on methodological details. The practical importance of a finding depends on personal circumstances, and the researcher's personal evaluation of that importance is outside the purview of science. Researchers are concerned with the accuracy of their conclusion, and given p(calculated), estimates of replicability add no information to this issue.

Note: Article Frick RW, SUNY Stony Brook, Dept Psychol, Stony Brook,NY 11790 USA

Keyword(s): confidence intervals; effect size; replication; statistical testing

Zurück | Weiter