Multialternative decision by sampling: A model of decision making constrained by process data.

Area of Science:

• Decision Science
• Cognitive Psychology
• Behavioral Economics

Background:

• Existing evidence accumulation models for multialternative choice lack consensus on the nature and process of evidence accumulation.
• This ambiguity limits the explanatory power of current decision-making models.

Purpose of the Study:

• To propose a novel model, multialternative decision by sampling (MDbS), that refines evidence accumulation processes in multialternative choice.
• To provide a quantitative account of choice phenomena, including attraction, compromise, and similarity effects, using process-tracing data.

Main Methods:

• Utilizing findings from process-tracing studies to constrain the evidence accumulation process.
• Extending the existing decision by sampling model to develop the MDbS model.
• Defining accumulated evidence as outcomes of pairwise ordinal comparisons between attribute values.

Main Results:

• The MDbS model quantitatively accounts for attraction, compromise, and similarity effects.
• MDbS demonstrates a broader empirical scope compared to existing models in explaining choice phenomena.
• The model's evidence accumulation mechanism is directly informed by empirical process-tracing data.

Conclusions:

• The MDbS model offers a more precise and empirically grounded framework for understanding multialternative choice.
• This approach advances the field by clarifying the nature of evidence accumulation in complex decision-making.
• The model's success suggests the utility of integrating process-tracing insights into computational models of choice.