The interpretation and use of statistics and probability appears to contain a variety of fallacies that allow the misuse of these tools. Has anyone come across a list of fallacies specific to these disciplines?

So far, this is what I've found online: https://en.wikipedia.org/wiki/Category:Probability_fallacies

  • 2
    I would be interested. There is a whole book called "Lying with statistics", but not necessarily about fallacies per se. Maybe the most common fallacy is to pick and choose the numbers one likes, while not showing the numbers they don't like.
    – Frank
    Commented May 6, 2023 at 2:26
  • 1
    @Frank So far, I have found this: en.wikipedia.org/wiki/Category:Probability_fallacies
    – user64314
    Commented May 6, 2023 at 2:51
  • Multiple such lists come up on simple googling, e.g. Misuse of statistics, Probabilistic Fallacies.
    – Conifold
    Commented May 6, 2023 at 6:08

1 Answer 1

  • Base rate fallacy
  • Confusion of the converse (or inverse)
  • Misuse of significance tests and confidence intervals
  • Assumption of independence
  • Conjunction fallacy
  • Data mining / data dredging / overfitting
  • Overreliance on averages and other summary metrics
  • Cherry picking of data
  • Confusing correlation with causation / false cause
  • Confusing an unconditional probability with a conditional probability and/or a marginal probability
  • Misuse of regression to the mean
  • Gambler's fallacy
  • Sampling bias
  • Berkson's bias
  • Survivor bias
  • Publication bias / newsworthiness bias / selective reporting
  • Overaggregation of data / overgenerlisation / Simpson's paradox
  • False extrapolation and/or interpolation
  • Observer effect
  • Texas sharpshooter fallacy
  • Poor quality data
  • Inadequate sample size
  • Incomparable or mixed or poorly understood categories
  • Poorly understood models and/or assumptions

You must log in to answer this question.