Feel the power: sample size and meaningful effects

There is, as the old saying goes, strength in numbers. We can look to numbers or statistics to help us interpret the evidence. Statistics can be used, for instance, to illustrate the safety and efficacy of interventions, to help establish treatment pathways, patients' satisfaction, and so on. However, we can be blinded by numbers; we can, according to the old trope, prove anything with statistics. Therefore, and at the risk of overstretching several metaphors, we must use statistics responsibly, otherwise there is the real risk that we end up like the Scottish poet Andrew Lang's (1844–1912) proverbial politician, using statistics like a drunk uses a lamp-post, in other words: to support rather than illuminate our arguments.

https://ift.tt/2vzqcIg