6P.3—Experimental design and method evaluation

Validity asks whether the method measures the intended variable; reliability asks whether repeats agree; precision concerns spread and measurement resolution. An evaluation should identify the limitation, its effect and a targeted improvement.

A colorimeter trace with close repeats may be reliable but invalid if the indicator is also affected by temperature. Add a control or redesign the measurement rather than simply taking more repeats.

More data do not repair a biased method, and an anomalous point is not automatically discarded.