RCTs for quantitative social researchers

Randomised controlled trials (RCTs) are heralded as the gold standard of research design in the social sciences. RCT principles are used in research at all levels of complexity from evaluating national social policies to experimenting with the impact of website designs (there often known as A/B testing).

This course is for social researchers who have a firm grasp of the foundations of quantitative research methods (e.g., linear regression and confidence intervals) and would like to learn how to design and analyse randomised controlled trials. The course incorporates a blend of presentations and participatory sessions, using examples from the social sciences.

The course covers:

• The fundamental problem of causal inference and how randomisation helps
• Neyman–Rubin’s potential outcomes framework
• An overview of RCTs, from idea and preregistration to final analysis and report
• Methods of randomisation
• Choosing a sample size
• Current best practice in analysing data from RCTs
• Managing “non-compliance” and other threats to study validity

By the end of the course participants will be able to:

• Explain the life cycle of an RCT
• Explain the benefits and costs of different RCT designs
• Use the potential outcomes framework to define what can be estimated in an RCT
• Choose a sample size for common designs
• Apply prior knowledge of quantitative analysis to RCTs
• Identify and address common threats to validity in RCTs