Academic journal article Psychonomic Bulletin & Review

Generalized Outcome-Based Strategy Classification: Comparing Deterministic and Probabilistic Choice Models

Academic journal article Psychonomic Bulletin & Review

Generalized Outcome-Based Strategy Classification: Comparing Deterministic and Probabilistic Choice Models

Article excerpt

Published online: 28 May 2014

© Psychonomic Society, Inc. 2014

Abstract Model comparisons are a vital tool for disentangling which of several strategies a decision makermay have used-that is, which cognitive processes may have governed observable choice behavior. However, previous methodological approaches have been limited to models (i.e., decision strategies) with deterministic choice rules. As such, psychologically plausible choice models-such as evidence-accumulation and connectionist models-that entail probabilistic choice predictions could not be considered appropriately. To overcome this limitation, we propose a generalization of Bröder and Schiffer's (Journal of Behavioral Decision Making, 19, 361-380, 2003) choice-based classification method, relying on (1) parametric order constraints in the multinomial processing tree framework to implement probabilistic models and (2) minimum description length for model comparison. The advantages of the generalized approach are demonstrated through recovery simulations and an experiment. In explaining previous methods and our generalization, we maintain a nontechnical focus-so as to provide a practical guide for comparing both deterministic and probabilistic choice models.

Keywords Judgment and decisionmaking . Model comparison . Strategy classification . Multinomial processing tree models . Minimumdescription length

(ProQuest: ... denotes formulae omitted.)

Introduction

One prominent viewpoint in judgment and decision making is based on the notion that individuals have various decision strategies at their disposal and that they will-more or less deliberately-select one or several for any given task and environment (e.g., Beach & Mitchell, 1978; Gigerenzer & Selten, 2001; Payne, Bettman, & Johnson, 1993; Weber & Johnson, 2009). More generally speaking, different underlying cognitive processes might govern observable judgments and decisions, and it is therefore vital to somehow identify these processes. To the degree that methods are capable of pinpointing how judgments and decisions are made, progress can be made in identifying determinants and bounding conditions of certain models or strategies-for example, in terms of the influence of different environmental structures (Bröder, 2003; Rieskamp & Otto, 2006), varying degrees of time pressure (Glöckner & Betsch, 2008c; Hilbig, Erdfelder, & Pohl, 2012; Payne, Bettman, & Luce, 1996; Rieskamp & Hoffrage, 2008), monetary information costs (Bröder, 2000; Newell & Shanks, 2003; Newell, Weston, & Shanks, 2003), different learning tasks and information formats (Bröder, Newell, & Platzer, 2010; Bröder & Schiffer, 2006; Pachur & Olsson, 2012; Söllner, Bröder, & Hilbig, 2013), the amount versus consistency of evidence (Glöckner & Betsch, 2012), and many more.

Despite the many extant investigations and important findings, identifying underlying decision strategies-or, more generally speaking, comparing process models of decision making-remains a challenge. Indeed "Behavioral Decision Research . . . is plagued with the problem of drawing inferences from behavioral data on cognitive strategies" (Bröder & Schiffer, 2003, p. 193). One approach is to focus on patterns of information acquisition (for an overview, see Schulte- Mecklenbeck, Kuhberger, & Ranyard, 2011) to infer which decision strategies were more or less likely to be applied (Glöckner, Fiedler, Hochman, Ayal, & Hilbig, 2012; Glöckner & Herbold, 2011; Johnson, Schulte-Mecklenbeck, & Willemsen, 2008; Payne, Bettman, & Johnson, 1988). However, information acquisition is not equivalent to information integration (Glöckner & Betsch, 2008c). For instance, a decision maker may search through all the information available but then integrate only a small subset of it. Thus, as a more common approach, the degree to which choice data are aligned with choice models or strategies is taken as an indicator of strategy use. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.