【摘要】 <正> 我们可以利用病例对照配对研究的方法控制某些混淆因素对所研究因素的影响。但是我们常用的2×2表的配对分析方法具有很大的局限性。如果想按多暴露水平分析,往往受到方法的限制而采取成组资料的分析方法,这样就更多还原

【Abstract】 A method of weighted least squares is introduced for the linear logistic analysis of matched pair case-control data.This method may be used to stratified analysis in matched pair multiple exposure level casecontrol study.As an example,it is used to risk estimates for stratified matched pair data by fohr exposure levels.Odds ratios and their confidence limits of comparing diffeerut expoure levels with reference group and statistics for hypothesis testing can be computed. This approach is limited in categorical risk variables, but continuous risk variables changed to categorical risk variables can also be used.更多还原