Randomised Controlled Trials: What, why and how? Pam Hanley 22 March 2013
What are RCTs? “Gold standard” for a trial – find out which intervention is most effective at achieving a specific, measurable outcome Most familiar from medicine eg drugs trials Divide a group of people, schools etc at random Compare “intervention” group(s) with a “control” group
Why do we need them? “… to test the effectiveness of new and existing interventions …; to learn what is working and what is not; and to adapt our policies so that they steadily improve and evolve both in terms of quality and effectiveness.” Haynes, Service, Goldacre & Torgerson (2012) (Cabinet Office Report: Test, Learn, Adapt)
Examples of RCTs in Education Reading eg beginning reading, struggling readers Maths Technology and reading, maths etc Primary science School reform programmes Social and emotional learning
Benefits to RCT participants Intervention groups: early triallists and some influence on development Control group: benefit from refinements made as a result of trial Participants are equally vital to the study regardless of group
Evaluation process Intervention A Control Initial sample Randomly assign to group Assess for eligibility and invite to participate Measure outcomes Intervention B
Challenges Recruiting schools Large-scale study Long-term commitment (legacy year) Changes in school staff (and schools)
Three possible objections to RCTs 1. Ethics: only unethical if we already know the intervention works 2. Education is too complex: but unknown factors are more likely to be balanced out through randomisation 3. RCTs don’t explain HOW: need to use in conjunction with other methods, eg observations, focus groups, interviews