HARKing

What is HARKing?

HARKing means Hypothesizing After the Results are Known: developing a hypothesis after seeing the data, then presenting it as a prediction made beforehand. It makes an explanation fitted to a result look like a successful test of a prior idea.

How HARKing works

Imagine a researcher who measures 15 outcomes and finds one significant effect. After seeing it, the researcher develops an explanation for why that particular outcome should change, then writes the paper as though this explanation motivated the study. That is HARKing. If the paper also hides the other 14 outcomes, it adds selective reporting.

The two practices are related, but they are not identical. HARKing concerns the false history of a hypothesis. Selective reporting concerns which results readers get to see. Norbert Kerr’s original 1998 paper describes forms of HARKing and examines its incentives and costs.

HARKing and False Positives

A chance pattern can look more convincing when it is presented as a prediction made in advance. HARKing hides that the result helped shape the hypothesis. In the example above, testing many outcomes and selectively reporting one also affects how the evidence should be judged.

Why it matters

Exploration is a useful part of science. A surprising result can suggest an excellent new question. The problem begins when the paper erases that sequence. Predicting an outcome before seeing it and explaining it afterward are different achievements, and readers need to know which one occurred.

Preregistration creates a dated record of planned hypotheses and analyses. Clear reporting can then distinguish those plans from ideas developed later. Neither prevents researchers from learning something unexpected; both help keep the account honest.