HARKing

HARKing, short for hypothesizing after the results are known, is a term the social psychologist Norbert Kerr coined for the research practice of presenting a hypothesis formed after seeing a study's results in the introduction of a paper as though it had been proposed beforehand. It can take the form of dropping an original hypothesis that failed to pan out and silently replacing it with one invented to fit the data that were actually found, or of retrieving an existing hypothesis from the published literature after the fact and presenting it as though it had guided the study from the start. Kerr traced the practice to incentives built into academic publishing, which favors papers built around hypotheses confirmed in advance and tends to reject reports of null findings, pressures that make a study look more convincing when its hypothesis appears to have predicted the result rather than explained it after the fact.