Experimental control
Also: control condition, baseline
A comparison condition kept unchanged so an intervention's effect can be distinguished.
Controls expose background drift, confounds, and measurement artifacts by preserving all relevant conditions except the tested factor.
Example Measure the fan normally before repeating with the inlet restriction.
Common misconception A control is not necessarily a separate device or a feedback controller.