NotesFAQContact Us
Search Tips
ERIC Number: ED126639
Record Type: RIE
Publication Date: 1976-Apr
Pages: 12
Abstractor: N/A
Statistical Power Analysis of Research in "Exceptional Children."
Schmelkin, Liora
A statistical power analysis of research studies in special education reported in volumes 39 and 40 of Exceptional Children (1972-73 and 1973-74) was conducted. The basic concepts of a power analysis (Type I and Type II errors, and conventional effect sizes) were reviewed, and the studies evaluated for statistical power. Using the .05 level of significance, the average power to detect small effects was .11 (5% having a better than 50-50 chance to declare significant findings), for medium effects the average power was .49 (43% having a better than 50-50 chance), and for large effects the average power was .82 (76% having a better than 50-50 chance). There are several approaches to increasing power, including increasing the number of Ss, increasing the level of alpha, altering the research design, using directional tests, and using highly reliable measures. Statistical power analysis should become a part of research training and one of the criteria for the evaluation of research reports. (Author/IM)
Publication Type: Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: N/A
Sponsor: N/A
Authoring Institution: N/A
Note: Paper presented at the Annual International Convention, The Council for Exceptional Children (54th, Chicago, Illinois, April 4-9, 1976)