Understanding statistical significance

Robust evaluation means assessing whether programmes have made a difference over and above what would have happened without them, which is known as the ‘counterfactual’. The best way to estimate this difference is to find a control or comparison group, which is how the Data Labs model works.

This short guide is designed as an introduction to the field, explaining the key terminology to non-specialists. The paper covers comparison groups and how they are used to benchmark your outcomes. It explains what statistical significance testing means, and how to interpret results—especially when your outcome is inconclusive.