On Logit Confidence Intervals for the Odds Ratio with Small Samples
Author:
- 1 June 1999
- journal article
- research article
- Published by Oxford University Press (OUP) in Biometrics
- Vol. 55 (2) , 597-602
Abstract
Summary.Unless the true association is very strong, simple large‐sample confidence intervals for the odds ratio based on the delta method perform well even for small samples. Such intervals include the Woolf logit interval and the related Gart interval based on adding .5 before computing the log odds ratio estimate and its standard error. The Gart interval smooths the observed counts toward the model of equiprobability, but one obtains better coverage probabilities by smoothing toward the independence model and by extending the interval in the appropriate direction when a cell count is zero.Keywords
This publication has 11 references indexed in Scilit:
- Approximate Is Better than "Exact" for Interval Estimation of Binomial ProportionsThe American Statistician, 1998
- Comparison of Exact, Mid-p, and Mantel-Haenszel Confidence Intervals for the Common Odds Ratio across Several 2 × 2 Contingency TablesThe American Statistician, 1992
- Estimating the Variance of Empirical Logits and Contrasts in Empirical Log Probabilities1984
- The Comparison of Proportions: A Review of Significance Tests, Confidence Intervals and Adjustments for StratificationRevue de l'Institut International de Statistique / Review of the International Statistical Institute, 1971
- Alternative Analyses of Contingency TablesJournal of the Royal Statistical Society Series B: Statistical Methodology, 1966
- CONFIDENCE LIMITS FOR A CROSS‐PRODUCT RATIOAustralian Journal of Statistics, 1962
- On estimating binomial response relationsBiometrika, 1956
- THE ESTIMATION AND SIGNIFICANCE OF THE LOGARITHM OF A RATIO OF FREQUENCIESAnnals of Human Genetics, 1956
- ON ESTIMATING THE RELATION BETWEEN BLOOD GROUP AND DISEASEAnnals of Human Genetics, 1955
- On the Problem of Confidence IntervalsThe Annals of Mathematical Statistics, 1935