“All scientific work is incomplete, whether it be observational or experimental. All scientific work is liable to be upset or modified by advancing knowledge. That does not confer upon us a freedom to ignore the knowledge we already have, or to postpone the action that it appears to demand at a given time. Who knows, asked Robert Browning, but the world may end tonight? True, but on available evidence most of us make ready to commute on 8:30 the next day.” Sir Austin Bradford Hill, as quoted in his landmark 1965 paper on causation.
This is a nice example of mining the existing medical literature to discover trends and attitudes about research. The authors take too dim a view, in my opinion, of what observational studies can show, but the statistics are still interesting to follow. Continue reading
The Framingham Heart Study, a massive longitudinal observational study, has helped identify most of the risk factors for heart disease that we commonly accept today. This paper provides a historical overview of the study and the many valuable insights that it provided. Continue reading
Some asked a question about a retrospective study where you have a control cohort matched to a case cohort so the cohorts are similar on important (potentially confounding) variables. I pointed out that the two consecutive words “case cohort” are ambiguous and tried to explain how I define a retrospective cohort design versus a (retrospective) case-control design. Continue reading
Someone posted a question noting that most of the statistical consulting projects that they worked on finished in a reasonable time frame, a few were outliers. They took a lot longer and required a lot more effort by the statisticians. Were there any common features to these outliers they wondered. So they asked if anyone else had identified methodological features of projects that went overtime. I only had a subjective impression, but thought it was still worth sharing. Continue reading
I’m working on a project where the researchers need a case-control study, though they may not know that yet. I want to show them what a case-control study gives them that would not be available with other methods. But I need to come up with a reasonable control group for the case-control design. It doesn’t have to be perfect, but it can’t be a totally stupid control group either. This article is the classic reference on the theoretical principles that underlie the selection of controls in a case-control study. Continue reading
Dear Professor Mean, I review a lot of observational studies in the literature, and I am concerned about the response rates and when they fall so low that they tend to produce problems with selection bias. I’ve heard that anything lower than 80% is a problem. Is that correct? Continue reading
This video gives a non-technical overview of the strengths and weaknesses of observational studies. Continue reading