Without randomisation you might accidentally end up comparing two completely different groups of people – apples with pears, if you like. For example, one group may have far more smokers than the other. Smoking can affect how a treatment works.
Studies that compare two groups without randomising them are called ‘observational studies’. These studies provide much less reliable evidence than clinical trials. They are useful for other types of research and can help researchers to formulate hypotheses to be tested in controlled experiments.
In some trials, to avoid systematic error and bias, neither people taking part nor their doctors, know if they are receiving the real treatment being tested or a fake treatment that exactly resembles the real thing (placebo). These trials are referred to as ‘double-blinded’.
Meanwhile other trials are ‘blinded’ to avoid systematic error or bias. This means that the scientists and doctors observing and analysing the trial data do not know which treatment participants were assigned to.