From the course: Meta-analysis for Data Science and Business Analytics
Unlock the full course today
Join today to access over 22,600 courses taught by industry experts or purchase this course individually.
Logarithms in risk and odds ratios
From the course: Meta-analysis for Data Science and Business Analytics
Logarithms in risk and odds ratios
- [Instructor] When you read about either risk ratios or odds ratios in a meta-analysis, you sometimes see that the authors work with the logarithms, or the logs, of those ratios. The same is true in logistic regression, where the log of the odds, or logit, have certain characteristics that are very useful as compared to either the raw probabilities or the odds. The rational for using logs in a meta-analysis of studies that report binary outcomes is weaker than the rational in logistic regression, but you're likely to encounter it, so you may as well understand it. The accepted method is to convert risk ratios or odds ratios to logarithms before you combine them in a meta-analysis. Confidence intervals around the ratios for an individual study, or around the ratio for a full meta-analysis, use the same logarithmic approach. There are several reasons for this. One reason is that the odds ratio's distribution is positively skewed, but the logs of the odds ratios have a distribution…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.