What does a control do?

MBG Türkiye / journal

A measurement on its own tells us surprisingly little. To understand whether a change is meaningful, researchers need something appropriate to compare it with.

A negative control helps identify background effects; a positive control checks that a procedure can detect the phenomenon it is designed to measure. Neither is a universal recipe. The right comparison depends on the question.

When reading a scientific report, ask which alternatives the controls rule out. That is often the shortest path to understanding what the experiment really shows.