Practical problems often make it difficult to obtain homogeneous and similar samples. For example, samples may involve individuals of different ages and may have to be taken on different days of the week. Individuals differ in numerous ways, and real effects can arise on different days. It could be said, quite rightly, that samples differ because a variety of effects are always present, each creating a difference. In other words, no matter how we aim to obtain homogeneous samples, we will end up with multiple effects. In the past, when analysis involved lengthy procedures, this was a nuisance. Now, with the availability of computer packages that provide rapid and more versatile processing, multivariate data analysis is seen to be a great advantage and has in many areas taken over from the simplistic methods I have been describing.