The technique of ranking is similar to that of rating, except that respondents are asked to compare a series of items, rather than place them on an absolute scale. Thus, for example, respondents might be presented with a series of occupational titles, and asked to rank these in terms of their general desirability, perceived social standing, suitability as employment for a woman, or whatever. The most common method of ranking uses the technique of paired comparisons, in which a series of paired stimuli (for example two different methods of allocating a reward) are presented to respondents, who are asked to say which is the greater or lesser in terms of some attribute in which the researcher is interested (in this case, perhaps, fairness). Each item is paired against every other item, and a method of scoring is used to construct a scale-value for the complete range of items, perhaps by simply scoring 1 for every preference recorded. (However, this would be a fairly crude technique, since it assumes that there is an equal interval between an item chosen twice and another chosen three times: various techniques have therefore been devised for deriving more complex and precise scale-values.) The method of equal appearing intervals developed by L. L. Thurstone is generally treated as a ranking-scale procedure.