Linear regression models using meta regression
The Meta Regression procedure builds linear regression models between the effect size and the covariate. The target or the dependent variable serves as the effect size; the predictors are the covariates, which are also called moderators and recorded as the study level (for example, the study location, study test environment and drug administration method). The goal of meta-regression is to explore and explain the between-study heterogeneity (difference in the true effect size of the individual studies) as a function of moderators.
Research studies involving type II diabetes medication
Several research studies were conducted to investigate a faddish but debatable medicine to help treat type II diabetes. The oral medicine was claimed to be able to reduce the blood glucose level after meals. Data were collected from different research sites from 1979 to 1986.
An analyst wants to draw statistical inference about the effect of the oral medicine by computing the effect size, log risk-ratio, and corresponding variance. The analyst would like to know the relationship between the effect size of log risk-ratio and the moderator (length (length of applying the medicine in days)), and would also like to explore to what extent the between-study heterogeneity can be accounted for based on the moderator.
The original data are collected in the glucose.sav sample file The data from glucose.sav was cleaned and stored in the glucose_length.sav sample file. See the topic Sample Files for more information.