Think of things like "lines of code written" or "bugs closed" as measurements of productivity or quality. These are real things that people have used in real published studies - and any conclusion drawn from them is obviously bogus.

Data gathered from a poorly designed study? I don't think it needs explaining why making decisions on nonsense is just that.

Measurement noise, other experimental errors, small sample size, cherry picking. It's worse than no data because you can draw incorrect conclusions.

Have a really dumb example:

All of the developers I work with are short, therefore all developers must be short, and we could save time by filtering job applicants by height.

(Note: I work with only one developer. He's great!)