Post-designation

Post-designation (also fishing for data) is a logical fallacy that occurs when a conclusion is drawn from correlations observed in a sample, but only after collection of the sample and without declaring in advance what correlations the experimenter was expecting to find. This fallacy thus circumvents the scientific method of having a pre-designated experimental hypothesis and testing for the null hypothesis or worrying about statistical significance before looking at the data.

Cogito ergo sum
Logic and rhetoric
Key articles
General logic
Bad logic
v - t - e
Not to be confused with a statistical "fishing expedition" (abusing p-values) or "cherry picking" (selecting a biased sample).

The fallacy is a circular fallacy, since any given data will always support some conclusion. Because we were looking for anything, we are bound to find it.

Examples
  • In looking at the difference between 100 Christians and 100 atheists, we found that Christians were significantly more likely to eat tuna fish.
  • Three of my four children were born in February, and all three were left-handed. Apparently most people born in February are left-handed.
  • We took a survey of our class and discovered that, out of 30 students, seven were born in January. We conclude that college students are much more likely to be born in January than in any other month.

Legitimate use

Using peculiarities in a sample to suggest new lines of research is not a bad idea, but requires going through the scientific process all over again. When this is done, the sample is the minor premise of a retroduction rather than the major premise of an induction.

gollark: Hmm.
gollark: "Smart contracts as applied to curses"
gollark: "Precognitive processing: could it be the next technology to revolutionize communication?"
gollark: "0.7% of Internet-connected computers are now posessed by ghosts."
gollark: And protestors insisting that cursing random bacteria to store data was "uNnAtUrAl".

See also
This article is issued from Rationalwiki. The text is licensed under Creative Commons - Attribution - Sharealike. Additional terms may apply for the media files.