Logical fallacies

correlation is not causation

Two things move together. That is where your evidence stops.

  • “Teams using this tool ship faster, so adopt it.”

    “Do fast teams adopt it, or does it make teams fast?”

    The arrow may point the other way. Good teams pick good tools.

Four things can produce a correlation: A causes B, B causes A, something else causes both, or coincidence. The data alone can’t tell you which, and picking the flattering one isn’t analysis.

How to answer it: ask for the mechanism, or for what happened when someone changed A on purpose. That’s the whole reason controlled experiments exist.

Now go look for the mechanism.