Scientific Rigor: Foundations That Matter
HARKing—Hypothesizing After Results are Known—happens when rexploratory findings are described as planned hypotheses. A pattern spotted in the data, an unexpected correlation, a statistical anomaly gets reframed as a confirmed discovery. The result looks more significant than it is, and what amounts to exploratory analysis gets published as confirmatory testing.
This matters because it contributes to research findings that don’t hold up. The scientific literature accumulates false positives. Subsequent researchers waste time and resources trying to build on discoveries that were, in essence, statistical accidents. The problem isn’t usually intentional; it’s a natural human tendency to make sense of patterns once we’ve seen them, and to present our work in the most compelling light.
The solution is structural. Pre-registration changes the dynamic. When researchers publicly document their hypotheses, sample sizes, and analytical plans before collecting data, they create a transparent record. There’s no way to rewrite history; the plan is already public. This protects both the researcher and the field.
Scientific journals can help too. Making space for exploratory findings—labeling them as such—means researchers can report what the data actually showed, even if it differs from the original plan. “We found this, though we didn’t anticipate it” is honest and valuable. Distinguishing clearly between confirmatory testing (which builds confidence in a theory) and exploratory analysis (which generates questions for future work) removes the incentive to blur the line.
When this shift happens, published conclusions become more trustworthy. Science advances more efficiently because follow-up research is built on genuine discoveries, not statistical flukes.
