In each of two experiments respondents were asked to narrow a given set of multiattribute options to a final choice in three successive stages. Some participants were asked to record which options they would include at each stage while others were asked to record which options they would exclude at each stage. The choice options in Experiment 1 were 24 alternative automobile descriptions, while in Experiment 2 participants were presented with 32 different graduate schools. Consistent with the existence of separate criteria for inclusion and exclusion, significantly more options were retained prior to final choice (in Stages 1 and 2) under exclusion than under inclusion. This finding held both for options with and without missing information. Nevertheless, when we examined Stage 3 in each experiment, we found no systematic differences between the distribution of final choices in each condition. These results, along with supplemental data, support a theoretical account in which options about which the decision maker is unsure are retained under exclusion but not under inclusion. © 1998 John Wiley & Sons, Ltd.
[1]
Richard E. Potter,et al.
Decision Making When the Acceptable Options Become Unavailable
,
1994
.
[2]
Gregory B. Northcraft,et al.
Decision bias and personnel selection strategies
,
1987
.
[3]
Irwin P. Levin,et al.
Individual differences in dealing with incomplete information: Judging clinical competence
,
1991
.
[4]
K. A. Ericsson,et al.
Protocol analysis: Verbal reports as data, Rev. ed.
,
1993
.
[5]
Irwin P. Levin,et al.
Phased Narrowing: A New Process Tracing Method for Decision Making
,
1995
.
[6]
Richard D. Johnson,et al.
Making Decisions With Incomplete Information: the First Complete Test of the Inference Model
,
1989
.