NotesFAQContact Us
Search Tips
ERIC Number: ED109146
Record Type: Non-Journal
Publication Date: 1975-Apr
Pages: 26
Abstractor: N/A
Reference Count: N/A
Small Sample Comparisons of the Cochran Q and the Minimum X sub one squared Statistics.
Vitaliano, Peter Paul
The Cochran Q and the Minimum X sub one squared statistics are two ways to test a hypothesis of equivalent correlated proportions. This study investigated the small sample properties of Q and X sub one squared by Monte Carlo methods. The observed distributions were compared for their rates of covergence to the limiting theoretical X sub one squared distribution, and for the degree to which their error rates approximated the nominal error rates. These latter comparisons allowed for idiosyncrasies that exist in the Q test of the correlated proportion hypothesis. Results show that the X sub one squared statistics is more powerful than the Q statistic for testing hypotheses of equivalent correlated proportions. (Author)
Publication Type: Reports - Research
Education Level: N/A
Audience: N/A
Language: N/A
Sponsor: N/A
Authoring Institution: N/A