Estimation of attribute weights in a multiattribute choice situation

Multiattribute judgement processes have regularly been investigated by means of structural modelling techniques. With these techniques mathematical models are fitted on the data and attribute weights reflecting the relative importance of the attributes during the process can be derived. Since many observations are needed for fitting such models, subjects are mostly requested to judge a large number of alternatives. In this article it is demonstrated how in situations in which the subject has to select the most attractive alternative out of a set, the decision process can also be studied by fitting a mathematical model, The validity of the estimates of the attribute weights provided by this method are investigated in an empirical and a simulation study, It appears that this method gives valuable information about the process, but that the estimates of the weights are slightly biased.