Contrasting holistic-compensatory with probabilistic heuristic strategies in multi-attribute decisions

Area of Science:

• Cognitive Psychology
• Decision Science
• Behavioral Economics

Background:

• Traditional views suggested humans use simplified strategies (e.g., Take The Best - TTB) under time pressure in multi-attribute decision-making.
• Recent research challenges this, proposing more complex strategies may be employed even with limited information.
• The study investigates a probabilistic extension of TTB, the Single Probabilistic Attribute (SPA) model, as an alternative heuristic.

Purpose of the Study:

• To test the validity of the Single Probabilistic Attribute (SPA) model against existing decision-making models, including TTB and the normative weighted-average (WAV) model.
• To differentiate between heuristic (SPA/gTTB) and compensatory (WAV) models using decision times.
• To analyze individual differences in decision strategies under speeded conditions.

Main Methods:

• Computational comparison of SPA, gTTB, and WAV models using choice data.
• Empirical data collection via a speeded multi-attribute decision task with varying numbers of numerical attributes (3, 4, 5).
• Analysis of decision times and choice patterns from 117 participants across two experiments.

Main Results:

• The SPA model showed comparable or better fit than gTTB for 3-5 attributes.
• Decision time analysis provided strong evidence against the SPA model.
• The majority of participants (70%) exhibited choices consistent with the compensatory weighted-average (WAV) model, making fast decisions (<1.5s).

Conclusions:

• Contrary to some prior research, findings suggest that under open-view, speeded conditions, most individuals employ fast, compensatory decision-making strategies (WAV).
• A significant portion of participants (approx. 30%) did utilize TTB-like strategies.
• Decision time is a crucial differentiator, supporting normative models over simpler heuristic accounts in this context.