ABSTRACT
The example in this chapter illustrates a number of points arising in the analysis of 'before and after' studies, in skeleton form that investigations in which some property is measured on each individual before 'treatment', a treatment then applied and then the same property re-measured. It is common in such studies to have a control group receiving a placebo or dummy treatment, choice between treatment and placebo being randomized with due concealment. Where there is no control group, a comparison 'after' versus 'before' can be made, but the interpretation of any difference as attaching specifically to the treatment is to be made with some reserve. Consider the measurement of response the systolic blood pressure. This is certainly legitimate: in the present instance the change in a patient's blood pressure is of immediate relevance both as a basis for analysing the full set of data and as an index of 'success' for that individual.
