Statistical Methods in Medicine

Analysis of ResultsMeasurementSufficient reference has already been made to the importance of ensuring that any two compared measurements are made on individuals or groups that are meaningfully comparable. It must further be ensured that the measurements themselves are derived in comparable ways — in other words, in such ways that any differences reflect real differences between the things measured and not merely differences in the technics of measurement. For many objective measurements there is no great difficulty on this score, sex, age, temperature and the presence or absence of specific disease, for instance, being usually determinable with little . . .