Multialternative Decision Field Theory Research Articles

A Sequential Sampling Approach to the Integration of Habits and Goals

AbstractHabits often conflict with goal-directed behaviors and this phenomenon continues to attract interests from neuroscientists, experimental psychologists, and applied health psychologists. Recent computational models explain habit-goal conflicts as the competitions between two learning systems, arbitrated by a central unit. Based on recent research that combined reinforcement learning and sequential sampling, we show that habit-goal conflicts can be more parsimoniously explained by a dynamic integration of habit and goal values in a sequential sampling model, without any arbitration. A computational model was developed by extending the multialternative decision field theory with the assumptions that habits bias starting points of preference accumulation, and that goal importance and goal relevance determine sampling probabilities of goal-related attributes. Simulation studies demonstrated our approach’s ability to qualitatively reproduce important empirical findings from three paradigms – classic devaluation, devaluation with a concurrent schedule, and reversal learning, and to predict gradual changes in decision times. In addition, a parameter recovery exercise using approximate Bayesian computation showcased the possibility of fitting the model to empirical data in future research. Implications of our work for habit theories and applications are discussed.

A process-oriented probabilistic linguistic decision-making model with unknown attribute weights

As an effective tool to describe qualitative evaluations, probabilistic linguistic term set (PLTS) can identify the different preference degrees for the possible linguistic evaluations. For the multi-attribute decision making (MADM) problems based on the PLTSs, making decisions is not instantaneous behavior but needs some time to complete information processing. Considering the dynamic nature of decision-making behavior, this study aims to develop a process-oriented probabilistic linguistic decision-making framework. First, we introduce the parameters in the probabilistic linguistic multi-alternative decision field theory (PLMDFT) model. An improved decision rule for selecting the optimal alternative(s) is also presented. Then, a deviation entropy-based model is developed to determine attribute weights. Furthermore, we construct a probabilistic linguistic decision-making framework based on the PLMDFT and deviation entropy. Finally, the constructed framework is applied to solve emergency scheme selection problem. Some discussion and comparative analysis is complemented to demonstrate the validity of the proposed framework.

Response-time data provide critical constraints on dynamic models of multi-alternative, multi-attribute choice.

Understanding the cognitive processes involved in multi-alternative, multi-attribute choice is of interest to a wide range of fields including psychology, neuroscience, and economics. Prior investigations in this domain have relied primarily on choice data to compare different theories. Despite numerous such studies, results have largely been inconclusive. Our study uses state-of-the-art response-time modeling and data from 12 different experiments appearing in six different published studies to compare four previously proposed theories/models of these effects: multi-alternative decision field theory (MDFT), the leaky-competing accumulator (LCA), the multi-attribute linear ballistic accumulator (MLBA), and the associative accumulation model (AAM). All four models are, by design, dynamic process models and thus a comprehensive evaluation of their theoretical properties requires quantitative evaluation with both choice and response-time data. Our results show that response-time data is critical at distinguishing among these models and that using choice data alone can lead to inconclusive results for some datasets. In conclusion, we encourage future research to include response-time data in the evaluation of these models.

Reversing the similarity effect: The effect of presentation format

A context effect is a change in preference that occurs when alternatives are added to a choice set. Models of preferential choice that account for context effects largely assume a within-dimension comparison process. It has been shown, however, that the format in which a choice set is presented can influence comparison strategies. That is, a by-alternative or by-dimension grouping of the dimension values encourage within-alternative or within-dimension comparisons, respectively. For example, one classic context effect, the compromise effect, is strengthened by a by-dimension presentation format. Extrapolation from this result suggests that a second context effect, the similarity effect, will actually reverse when stimuli are presented in a by-dimension format. In the current study, we presented participants with a series of apartment choice sets designed to elicit the similarity effect, with either a by-alternative or by-dimension presentation format. Participants in the by-alternative condition demonstrated a standard similarity effect; however, participants in the by-dimension condition demonstrated a strong reverse similarity effect. The present data can be accounted for by Multialternative Decision Field Theory (MDFT) and the Multiattribute Linear Ballistic Accumulator (MLBA), but not Elimination by Aspects (EBA). Indeed, when some weak assumptions of within-dimension processes are met, MDFT and the MLBA predict the reverse similarity effect. These modeling results suggest that the similarity effect is governed by either forgetting and inhibition (MDFT), or attention to positive or negative differences (MLBA). These results demonstrate that flexibility in the comparison process needs to be incorporated into theories of preferential choice.

Multi-attribute, multi-alternative models of choice: Choice, reaction time, and process tracing

The first aim of this research is to compare computational models of multi-alternative, multi-attribute choice when attribute values are explicit. The choice predictions of utility (standard random utility & weighted valuation), heuristic (elimination-by-aspects, lexicographic, & maximum attribute value), and dynamic (multi-alternative decision field theory, MDFT, & a version of the multi-attribute linear ballistic accumulator, MLBA) models are contrasted on both preferential and risky choice data. Using both maximum likelihood and cross-validation fit measures on choice data, the utility and dynamic models are preferred over the heuristic models for risky choice, with a slight overall advantage for the MLBA for preferential choice. The response time predictions of these models (except the MDFT) are then tested. Although the MLBA accurately predicts response time distributions, it only weakly accounts for stimulus-level differences. The other models completely fail to account for stimulus-level differences. Process tracing measures, i.e., eye and mouse tracking, were also collected. None of the qualitative predictions of the models are completely supported by that data. These results suggest that the models may not appropriately represent the interaction of attention and preference formation. To overcome this potential shortcoming, the second aim of this research is to test preference-formation assumptions, independently of attention, by developing the models of attentional sampling (MAS) model family which incorporates the empirical gaze patterns into a sequential sampling framework. An MAS variant that includes attribute values, but only updates the currently viewed alternative and does not contrast values across alternatives, performs well in both experiments. Overall, the results support the dynamic models, but point to the need to incorporate a framework that more accurately reflects the relationship between attention and the preference-formation process.

Attraction Effect in Risky Choice Can Be Explained by Subjective Distance Between Choice Alternatives

Individuals make decisions under risk throughout daily life. Standard models of economic decision making typically assume that people evaluate choice options independently. There is, however, substantial evidence showing that this independence assumption is frequently violated in decision making without risk. The present study extends these findings to the domain of decision making under risk. To explain the independence violations, we adapted a sequential sampling model, namely Multialternative Decision Field Theory (MDFT), to decision making under risk and showed how this model can account for the observed preference shifts. MDFT not only better predicts choices compared with the standard Expected Utility Theory, but it also explains individual differences in the size of the observed context effect. Evidence in favor of the chosen option, as predicted by MDFT, was positively correlated with brain activity in the medial orbitofrontal cortex (mOFC) and negatively correlated with brain activity in the anterior insula (aINS). From a neuroscience perspective, the results of the present study show that specific brain regions, such as the mOFC and aINS, not only code the value or risk of a single choice option but also code the evidence in favor of the best option compared with other available choice options.

Rigorously testing multialternative decision field theory against random utility models.

Cognitive models of decision making aim to explain the process underlying observed choices. Here, we test a sequential sampling model of decision making, multialternative decision field theory (MDFT; Roe, Busemeyer, & Townsend, 2001), on empirical grounds and compare it against 2 established random utility models of choice: the probit and the logit model. Using a within-subject experimental design, participants in 2 studies repeatedly choose among sets of options (consumer products) described on several attributes. The results of Study 1 showed that all models predicted participants' choices equally well. In Study 2, in which the choice sets were explicitly designed to distinguish the models, MDFT had an advantage in predicting the observed choices. Study 2 further revealed the occurrence of multiple context effects within single participants, indicating an interdependent evaluation of choice options and correlations between different context effects. In sum, the results indicate that sequential sampling models can provide relevant insights into the cognitive process underlying preferential choices and thus can lead to better choice predictions.

Supplemental Material for Rigorously Testing Multialternative Decision Field Theory Against Random Utility Models

Supplemental Material for Rigorously Testing Multialternative Decision Field Theory Against Random Utility Models

The 2N-ary Choice Tree Model for N-Alternative Preferential Choice

The 2N-ary choice tree model accounts for response times and choice probabilities in multi-alternative preferential choice. It implements pairwise comparison of alternatives on weighted attributes into an information sampling process which, in turn, results in a preference process. The model provides expected choice probabilities and response time distributions in closed form for optional and fixed stopping times. The theoretical background of the 2N-ary choice tree model is explained in detail with focus on the transition probabilities that take into account constituents of human preferences such as expectations, emotions, or socially influenced attention. Then it is shown how the model accounts for several context-effects observed in human preferential choice like similarity, attraction, and compromise effects and how long it takes, on average, for the decision. The model is extended to deal with more than three choice alternatives. A short discussion on how the 2N-ary choice tree model differs from the multi-alternative decision field theory and the leaky competing accumulator model is provided.

Cognitive model based fashion style decision making

The decision of the style of a garment was affected by not only the physical attributes of the components of the garment but also the decision context. Lacking consideration of the decision context leads to the prediction failure of many models in intelligent fashion design systems. This paper proposed to integrate the cognitive model – Multi-alternative Decision Field Theory (MDFT) with Genetic Algorithm to tackle the context problem, specifically, the choice set when a decision was made. A unified S expression and its sign deciding mechanism was given out. Experiments showed that the proposed model gives out a better prediction of human fashion style decision than GA only model.

Prospect relativity: how choice options influence decision under risk.

In many theories of decision under risk (e.g., expected utility theory, rank-dependent utility theory, and prospect theory), the utility of a prospect is independent of other options in the choice set. The experiments presented here show a large effect of the available options, suggesting instead that prospects are valued relative to one another. The judged certainty equivalent for a prospect is strongly influenced by the options available. Similarly, the selection of a preferred prospect is strongly influenced by the prospects available. Alternative theories of decision under risk (e.g., the stochastic difference model, multialternative decision field theory, and range frequency theory), where prospects are valued relative to one another, can provide an account of these context effects.

Intuition Beyond Recognition: When Less Familiar Events Are Liked More

Intuition Beyond Recognition: When Less Familiar Events Are Liked More

Multialternative decision field theory: A dynamic connectionst model of decision making.

The authors interpret decision field theory (J. R. Busemeyer & J. T. Townsend, 1993) as a connectionist network and extend it to accommodate multialternative preferential choice situations. This article shows that the classic weighted additive utility model (see R. L. Keeney & H. Raiffa, 1976) and the classic Thurstone preferential choice model (see L. L. Thurstone, 1959) are special cases of this new multialternative decision field theory (MDFT), which also can emulate the search process of the popular elimination by aspects (EBA) model (see A. Tversky, 1969). The new theory is unique in its ability to explain several central empirical results found in the multialternative preference literature with a common set of principles. These empirical results include the similarity effect, the attraction effect, and the compromise effect, and the complex interactions among these three effects. The dynamic nature of the model also implies strong testable predictions concerning the moderating effect of time pressure on these three effects.

Multialternative decision field theory: A dynamic connectionst model of decision making.

Multialternative decision field theory: A dynamic connectionst model of decision making.