EURO 2025 Leeds
Abstract Submission

228. Use of heuristics in pairwise multicriteria choice tasks: An experimental study

Invited abstract in session MC-61: Advances in behavioral decision analysis 2, stream Behavioural OR.

Monday, 12:30-14:00
Room: Maurice Keyworth G.31

Authors (first author is the speaker)

1. Jyrki Wallenius
Information and Service Management, Aalto University School of Business
2. Eeva Vilkkumaa
Department of Information and Service Management, Aalto University, School of Business
3. Matias Kivikangas
Information and Service Management, Aalto University School of Business
4. Pekka Malo
Information and Service Economy, Aalto University School of Business
5. Kalyanmoy Deb
Computer Science and Engineering, Michigan State University
6. Konstantinos Katsikopoulos
Southampton Business School

Abstract

There is evidence that people use heuristics in decision-making that are somewhere between non-compensatory (such as Take The Best = TTB) and compensatory (such as tallying). Examples of such heuristics include the Take-Two heuristic, in which decisions are made based on two most salient pieces of information; and a threshold heuristic, in which information is gathered until the evidence supporting a given alternative exceeds some threshold.

Against this background, we study the following research questions:

1. What is the “fit” of different heuristics and their variations to available data?
2. How do these heuristics depend on the number of criteria?
3. How do these heuristics depend on the time spent on the task (i.e., learning)?

To address these questions, we use data from an experiment in which 48 subjects made altogether 5,670 pairwise choices between multicriteria decision alternatives (product bundles) so that the number of criteria ranged between 2-10. In addition to the subjects’ choices, we have gathered process data through questionnaires and eye-tracking measures. We use these data to study whether the outcomes as well as the process data align with TTB, tallying, Take-Two, and threshold heuristics. We also study, whether this alignment depends on the number of criteria and the time spent on the task. We propose novel ways of operationalizing Take-Two and threshold heuristics in the context of multicriteria decision making.

Keywords

Status: accepted

Back to the list of papers