Skip navigation


“Statistical prudence encourages the design of experiments [or studies] that allow conclusions to be drawn from them”

Source unknown

We’ve already covered the basic form of an experiment and an observational study, and how random allocation of the exposure or treatment (randomisation) can be used to minimise other sources of variation between groups and help us infer a causal association.  You have also been introduced to the idea of bias, this section revisits this more formerly.  

We cover the main features and strengths and weaknesses of different types of clinical trials (topic 9) and observational studies (topic 10). There is also a short overview of sample size and statistical power (topic 12) which is required when designing a study, and lectures on laboratory experiments (Topic 11) and Ethics (Topic 13).