ABSTRACT
How can statistical procedures be used to inform such a subjective process? The answer is that they can but it all depends on the types of question we ask and the types of evidence we use. Let’s assume we have two objects, ‘this’ and ‘that’ and two questions to answer, ‘Which is taller?’ and ‘Which is better?’ At the outset we can say that everyone should agree which of ‘this’ or ‘that’ is taller but we can have no certainty about judgements of which one is better. This, of course, is because my idea of what makes something better may be different from yours. I might prefer an apple to an orange but you might prefer an orange to an apple and who is to say that I am right and you wrong or vice versa? However, what if we were to deal not with individuals but with groups? Our question could then become, ‘How many think this is better than that?’ Having put the question in this way, we have created a situation in which everyone should agree, on the basis of the evidence, that more people prefer this or more that. We may not agree with the majority view but we have to accept that there is a majority view.
