We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
NOTE ON THE CHI-SQUARE STATISTIC OF ASSOCIATION IN 2 X 2 CONTINGENCY TABLES AND THE CORRECTION FOR CONTINUITY.
- Authors
Loosen, Franz
- Abstract
Research workers frequently do not have to make a decision but merely want some idea of how likely or unlikely the observed sample result is when some null hypothesis is true. Recently some authors argued that in those situations one can better report the smallest value of &agr; for which the hypothesis would have been rejected, rather than the sole fact of whether or not this hypothesis should be rejected at the traditional 0.05 of 0.01 levels of significance. In the present article the author wants to state that if one wishes merely to tell how unusual the sample result is as compared with the sampling distribution under the assumption that the hypothesis of independence is true, it is better to use a chi-square statistic which does not include a correction for continuity rather than a test statistic which does include such a correction, at least if one wishes to relate the chi-square statistic to the sampling distribution observed by repeated application of the whole sampling process. However, the author is not against a correction for continuity as such, according to him.
- Subjects
CHI-squared test; ANALYSIS of variance; CONTINUITY; STATISTICAL sampling; CONTINGENCY tables; DISTRIBUTION (Probability theory); STATISTICS
- Publication
Quality & Quantity, 1979, Vol 13, Issue 4, p351
- ISSN
0033-5177
- Publication type
Article
- DOI
10.1007/BF00188021