Au, W.T., and Wu, M. S., The Chinese University of Hong Kong, Hong Kong
There are no general rules specifying whether any group decision task should be structured as selection among discrete alternatives or as estimation of a quantitative judgment. A "naturally categorical" decision can be decided as a continuous judgment. This paper presents data on the effect of response scale [categorical vs. numerical] on group decision making. Student participants role-played members in a student admissions committee. Their job was to determine which candidates to be selected to join an undergraduate program. The experiment entailed a 2 2 (Response Scale [Categorical vs. Numerical] Decision Criterion [Tough vs. Easy]) factorial design. In the CATEGORICAL conditions, the group had to make a binary decision (i.e., select vs. not-select) for each of the 20 applicants. In the NUMERICAL conditions, the group had to base their binary decisions on numerical ratings of the merits of the candidates. In the TOUGH criterion condition, the chosen-applicant would be granted a conditional offer. In the EASY criterion condition, the chosen-applicant would be given an interview for further consideration. The results show that in the NUMERICAL conditions, more applicants were selected in the EASY criterion condition (M = 10.3) than in the TOUGH criterion condition (M = 6.5). However, the effect of Decision Criterion was negligible under the CATEGORICAL conditions (M's = 9.6 vs. 8.2). Compared to participants in the CATEGORICAL conditions, those in the NUMERICAL conditions (a) consulted other group members less frequently, (b) were more likely to dominate the discussion, (c) were more likely to agree with others' opinion, and (d) were less aware of their own mistakes.