Statement on statistical significance and uncertainty of impact estimates for EEF evaluations

This document discusses the role of uncertainty in the interpretation of EEF-funded evaluations and explains why the EEF does not use statistical significance to discuss findings. 

It builds on, and supersedes, the EEF’s note, Statistical uncertainty in randomised controlled trials, published in August 2018. 

This document has been developed in association with the EEF’s Evaluation Advisory Group, other experts, and some members of the EEF’s panel of evaluators. 

The EEF is grateful for all the feedback received.

Access document here:

  1. Updated: 6th February, 2020

    Statement on statistical significance and uncertainty of impact estimates for EEF evaluations