Replication
Also known as: Direct replication, Conceptual replication
Repeating a study to see whether its finding holds up.
What it means
Replication is the process of conducting a study again to determine whether its original result can be reproduced, the cornerstone of self-correction in science. A direct replication repeats the original methods as faithfully as possible to test the finding itself, whereas a conceptual replication tests the same hypothesis using different operationalizations to probe its generality. The two serve different functions and can mislead if confused: a failed conceptual replication may reflect a changed method rather than a false original, while only direct replication cleanly tests the specific claim. Systematic replication efforts exposed that many published effects are weaker or absent on repetition, catalyzing reforms across the behavioral sciences.
Examples
Several labs independently rerun a priming experiment with the original materials and fail to reproduce the published effect.
A team's landing-page test shows a 20% lift; rerunning the identical test the next month on fresh traffic shows nothing, and the 'win' quietly disappears from the roadmap.
A conceptual replication swaps the original's word puzzle for a video clip, so when it fails nobody can tell whether the effect is false or the new method simply differs.
First described in Foundational scientific norm; Open Science Collaboration (2015).
Where this comes up
- Adaptive Behavioral Interventions Through Design: A Case Study MethodologyDesign has a measurable influence on human behavior. Across workplaces, schools, public institutions, and urban environ…