Abstract
Responds to L. M. Hsu's proposal of a statistical procedure for testing the difference between effect sizes in independent studies. It is shown that the test proposed by Hsu is incorrect and that it tends to reject the null hypothesis of equal effect sizes too often. The statistical model underlying effect size estimation is clarified, and the exact distribution of the effect size estimator is given. The distribution is used to show that the usual size estimator is biased. (6 ref) (PsycINFO Database Record (c) 2006 APA, all rights reserved).
| Original language | English (US) |
|---|---|
| Pages (from-to) | 691-693 |
| Number of pages | 3 |
| Journal | Psychological Bulletin |
| Volume | 91 |
| Issue number | 3 |
| DOIs | |
| State | Published - May 1 1982 |
Fingerprint
Keywords
- statistical procedure for testing difference between effect sizes, criticism of L. M. Hsu's proposal
ASJC Scopus subject areas
- Psychology(all)
Cite this
}
Estimation and testing for differences in effect size : A comment on Hsu. / Hedges, Larry V.
In: Psychological Bulletin, Vol. 91, No. 3, 01.05.1982, p. 691-693.Research output: Contribution to journal › Article
TY - JOUR
T1 - Estimation and testing for differences in effect size
T2 - A comment on Hsu
AU - Hedges, Larry V.
PY - 1982/5/1
Y1 - 1982/5/1
N2 - Responds to L. M. Hsu's proposal of a statistical procedure for testing the difference between effect sizes in independent studies. It is shown that the test proposed by Hsu is incorrect and that it tends to reject the null hypothesis of equal effect sizes too often. The statistical model underlying effect size estimation is clarified, and the exact distribution of the effect size estimator is given. The distribution is used to show that the usual size estimator is biased. (6 ref) (PsycINFO Database Record (c) 2006 APA, all rights reserved).
AB - Responds to L. M. Hsu's proposal of a statistical procedure for testing the difference between effect sizes in independent studies. It is shown that the test proposed by Hsu is incorrect and that it tends to reject the null hypothesis of equal effect sizes too often. The statistical model underlying effect size estimation is clarified, and the exact distribution of the effect size estimator is given. The distribution is used to show that the usual size estimator is biased. (6 ref) (PsycINFO Database Record (c) 2006 APA, all rights reserved).
KW - statistical procedure for testing difference between effect sizes, criticism of L. M. Hsu's proposal
UR - http://www.scopus.com/inward/record.url?scp=0010676199&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0010676199&partnerID=8YFLogxK
U2 - 10.1037/0033-2909.91.3.691
DO - 10.1037/0033-2909.91.3.691
M3 - Article
AN - SCOPUS:0010676199
VL - 91
SP - 691
EP - 693
JO - Psychological Bulletin
JF - Psychological Bulletin
SN - 0033-2909
IS - 3
ER -