Horses for courses

RCTs versus observational studies is a false dichotomy.

Today, we read of our long-standing supporter Myra’s despair at the politically motivated use of evidence to justify catastrophic decisions made on the spur of the moment.

We sympathise, but we must keep going. As long as we are free to express our views and share our vision of science, we shall continue with the support of folks like Myra.

What are we to make of the seemingly intractable dichotomy between randomised comparisons and observational studies, between trials and cohort studies? Predictive models are not evidence, so we shall dispense with those in what follows.

Tom’s great compatriot Galileo Galilei founded and first used the method we still use, with few modifications today. He observed in the heavens, formulated a theory and then tested it with all kinds of experiments. Galileo’s latest biographers doubted he actually did experiments from the leaning tower of Pisa, as by then he had been cautioned by the Inquisition. Perhaps he designed them in his head or tested them secre…