Psychometric modeling of consumer decisions in primary health care.

A psychometric technique, functional measurement, was used to measure the preferences of 77 persons for hypothetical health care providers identified by various levels of cost, travel time, and waiting time; responses were subjected to analysis of variance and graphic analysis to determine the functional relation by which respondents combined the three attributes in arriving at their preferences. Both modes of analysis suggested multiplicative decision models, but these yielded lower values of R2 than did linear models derived for comparison. Respondents were grouped by the proportion of response variance that was explained by each of the three attributes, and decision models were derived for each group. Discriminant analysis of socioeconomic characteristics of the respondents identified the variables that distinguished the groups and contributed to forming their preferences.