Measuring dimensions of belief by unidimensional unfolding (Q1329123)

This study shows that when carefully constructed sets of statements are used as stimuli, both the ordinal and quantitative components of Coombs' theory of unidimensional unfolding [\textit{C. H. Coombs}, A theory of data. (1964); Psychology and mathematics. (1983)]\ provide convincing explanations of pair comparison preference judgements. This provides evidence for the hypothesis that attitudinal differences are quantitative, an hypothesis that is often presumed true in pychological research but one that is rarely tested. If there are quantitative differences between attitudes towards the same object then this quantitative character must have a source within the logical and semantic structure of the statements used. Speculations have been made here as to what this source might be. The significance of this research into the foundations of attitude measurement is that it shows that an assumption common to techniques as varied as \textit{L. L. Thustone}'s [J. Abnormal Soc. Psychol. 31, 384-400 (1927)], \textit{R. Likert}'s [Arch. Psychol. 140, 5-53 (1932)]\ and \textit{G. Ekman}'s [Percept. Motor Skills 15, 3-9 (1962)]\ is not only testable via Coombs' theory of unidimensional unfolding but may, in fact, be more than a pious hope.