Statistical autopsy

From R. A. Fisher, 1938:

To consult the statistician after an experiment is finished is often merely to ask him to conduct a post mortem examination. He can perhaps say what the experiment died of.

3 thoughts on “Statistical autopsy

  1. An excellent observation. I’ve had too many people ask me for statistics help after the fact, and I’m not even an expert.

  2. This is so true. The worst is when a grad student comes to you and you realize that some experiments they designed for their thesis research were massively underpowered and/or poorly designed. It’s just so disheartening.

  3. Completely agree, I work as a clinical research consultant in public health/ epi/biostat and the amount of times physicians in particular (no offence I am becoming one haha) come to me and say I want to this ….. and I’m like but Why, Why this test, Why this variable? Most don’t even get the gist of basic measures of tendency. Then the train wrecks of oh please look at this and I’m like its a dead horse, you can recycle it for meat (another project ran correctly) or bury it and never let it see the light of day!
    #Biostats #Epidemiology #PublicHealth #ClinicalResearch

Comments are closed.