Significance tests have their place
Article Abstract:
Null-hypothesis significance tests (NHST) reveal the sufficiency of evidence needed to rely on the sign of the population effect. However, this can only occur when the two-valued logic is replaced with the three-alternative hypothesis tests. Confidence intervals, the proposed alternative to NHST, are fraught with the same flaws suffered by NHST. Unlike confidence intervals, NHST provide the probabilities of a Type III error and of replication.
Publication Name: Psychological Science
Subject: Psychology and mental health
ISSN: 0956-7976
Year: 1997
User Contributions:
Comment about this article or add new information about this topic:
Significance testing in psychological research: some persisting issues
Article Abstract:
Significance testing in psychological research is allegedly flawed. First, significance tests are claimed to be fraught with Type I errors, or improper rejections of null hypotheses. However, rates have been discovered to be well estimated despite some deviations. Critics also pinpoint the high rate of Type II errors, or acceptances of false null hypotheses. Null hypotheses are, however, viewed as constructions rather than propositions.
Publication Name: Psychological Science
Subject: Psychology and mental health
ISSN: 0956-7976
Year: 1997
User Contributions:
Comment about this article or add new information about this topic:
- Abstracts: Dissociative experiences in the general population. Dissociation and childhood trauma in psychologically disturbed adolescents
- Abstracts: Subclinical eating disorders in adolescent women: a test of the continuity hypothesis and its psychological correlates
- Abstracts: The California Verbal Learning Test in the detection of incomplete effort in neuropsychological evaluation. Updating to the WAIS-III and WMS-III: Considerations for research and clinical practice
- Abstracts: Physical and psychological representations of faces: evidence from morphing. Perceiving real-world viewpoint changes
- Abstracts: Self-corrected reaching in a three-location delayed-response search task