|| Checking for direct PDF access through Ovid
Sixty-one subjects with a wide range of formal music training were presented with ascending and descending C major scales, each followed by a single probe tone drawn from the 13-note chromatic scale. They were asked to decide whether the fit of the probe to the preceding context was good or poor. For each probe, the proportion of good-fit replies was considered as a measure of the degree to which the context established an expectancy (i.e., primed) for that probe. For each subject, the ratings of all 13×2 probes were considered as a profile of the internal representation of expectancies generated by the major scale for that subject. A cluster analysis based upon the individual profiles was used to group and classify subjects. This resulted in groups which could be described in terms of the relative emphasis on tonic triad, diatonic, and proximity relations as previously identified by Krumhansl and Shepard (1979), Correlations of the average rating with the average reaction time for each of the 13×2 probe tones for each group suggested a relation between the consistency of the ratings (either always good fit or always poor fit) and the reaction time: reaction times were faster when ratings were consistent and slower when ratings were inconsistent. It is concluded that a dichotomous rating procedure reveals expectancy profiles for the major scale context that are consistent with those described by Krumhansl and Shepard (1979) and that reaction time measures complement these ratings.