Abstract
A 4 × 3 factorial experiment was conducted in which 60 Ss perused briefly a list of occupations, holding in mind one of four types of scales, and then estimated the total number that would fall into each category of one of three types of scales. The perusal scales had three discriminating categories on one or the other side of the “average” category and a single nondiscriminating category on the opposite side or had three categories on each side of the “average” one; or no designated scale was kept in mind during perusal. For estimation, one of the two unbalanced scales or the balanced one was used. The results indicated that the expected distribution of judgments was not differentially affected by the type of scale used during perusal but was by the type used for estimation. The unbalanced scales produced distributions weighted in the direction of the side of the scale containing the three discriminating categories.