An eye fixation analysis of multialternative choice

Eye fixations were recorded while subjects chose their most preferred of six used cars. Fixation sequences of the form X-Y-X-... were used to identify pair comparisons. This assumption was validated by verbal protocols and by a comparison between X-Y-X and X-Y-X-Y-... patterns. The results of three experiments showed: (1) that the multialternative choice process was composed primarily of pair comparisons; (2) that evaluative processing took placeonly during these comparisons; (3) that strategies for selecting the pairs were based primarily on information processing convenience; (4) that the sequence of pair comparisons was not used by the subjects as a sequential elimination process; and (5)that subjects changed their strategies to adapt to different task environments. These results have implications for current choice models and for the use of eye fixations in other complex tasks.     

Limits to decision-making ability in direct multiattribute alternative evaluation

In recent years there has been considerable interest in how decision strategies in choice problems depend on such characteristics as the number of available criteria and the number of available alternatives. Along with multicriteria choice problems, classification tasks are treated where alternatives are classified into several decision (evaluative) classes. We believe that in classification problems one's ability in information processing is limited and depends on task complexity. In this paper we describe two experiments involving categorization of a large number of multiattribute alternatives into two or four evaluative classes. The results demonstrate that the subjects use different decision strategies and that there is a limit in human ability to implement these two tasks.     

Strategies for multiattribute binary choice

Based on eye-fixation patterns, strategies for multiattribute binary choice were classified as holistic (within an alternative) or dimensional (within an attribute across alternatives). In a task environment hospitable to both strategies, dimensional processing predominated. Even for alternatives like simple gambles, which require holistic computations, dimensional strategies were used as often as holistic ones. The dimensional strategies were augmented by two procedures that simplify the computations. These simplification procedures reduce cognitive effort at the cost of a relatively small increase in errors. However, for about half the subjects the use of these simplification procedures led to systematic violations of expected utility theory on certain choices. Both the preference for dimensional over holistic strategies and the adoption of simplifying procedures are compatible with the desire to reduce cognitive effort. We propose that strategies are selected to minimize the joint cost of errors and effort.     

Regulatory fit from attribute‐based versus alternative‐based processing in decision making

This paper discusses the fit between attribute-based versus alternative-based processing and regulatory focus, and its impact on decision outcome valuation. Attribute-based processing was found to occur more frequently under prevention focus, whereas alternative-based processing occurred more frequently under promotion focus. The fit between prevention/promotion focus and attribute-based/alternative-based processing was found to enhance satisfaction with choices and the perceived monetary value of chosen options. Moreover, the effect of fit on outcome valuation was found to be mediated by ease of processing. Finally, the effects of fit on ease of processing and outcome valuation disappeared when consumers first practiced to process information based on either attributes or alternatives.     

The Influence of Decision Aids on Choice Strategies: An Experimental Analysis of the Role of Cognitive Effort

This paper examines the role of computer-based decision aids in reducing cognitive effort and therefore influencing strategy selection. It extends and complements the works reported in the behavioral decision theory literature on the role of effort and accuracy in choice tasks. The central proposition of this paper is that specific features can be incorporated within a set of decision aids that will alter the effort required to implement a particular choice strategy relative to other strategies, and that this will influence strategy selection by the decision maker. In a laboratory experiment, using a repeated measures design, 48 subjects performed a preferential choice task using different decision aids. Subjects were given different levels of support to reduce the cognitive effort associated with different preferential choice strategies. In particular, the decision aids provided varying levels of support for the processing associated with either elimination by aspects or additive difference strategies. The study examined how information processing strategies were influenced by the decision aids. The results show that decision aids which reduce the effort associated with the elimination by aspects strategy induce behaviors associated with elimination by aspects. More importantly, there was an interaction effect caused by the distinct behavior of the group which was provided support for additive difference but not for elimination by aspects. This group processed the information about available choices in a way that is consistent with an additive difference strategy. These findings indicate that a decision aid can induce additive processing by altering the relative effort requirements of different choice strategies.     

跨期决策中的维度差异偏好：眼动证据

在跨期决策研究领域,虽然基于维度的跨期模型得到了一些源自结果检验和过程检验的证据支持,但此类模型所假设的维度间差异比较的心理过程尚缺乏直接的过程证据。本研究通过两个眼动实验,系统考察了相关眼动指标对维度差异偏好的预测效应。结果发现,根据基于维度的权衡模型可有效拟合出个体在跨期决策中的维度差异偏好,并且反应时、眼跳注视熵和静态注视熵等指标均与维度差异偏好负相关,而基于维度的注意分配与维度差异判断正相关。这些研究发现支持了本研究所提出的跨期眼动模型的相关假设,证实了维度差异偏好与跨期决策的认知加工过程之间的联系,为基于维度的跨期模型提供了更直接的过程证据,并为今后跨期决策的眼动模型发展指明了新方向。     

Postdecision evaluation of choice alternatives as a function of valence of alternatives,choice, and expected delay of choice consequences

The effects of valence of alternatives, choice, and expected delay of choice consequences on postdecision evaluations of choice alternatives were studied by either giving subjects a choice between two experimental tasks (both either pleasant or unpleasant) or assigning subjects their initially preferred task. Crosscutting valence and choice, subjects expected either no delay, 10-min delay, or 30-min delay before engaging in the task. All subjects evaluated both tasks immediately after the delay instructions but before experiencing the actual delay. Results confirmed the expected dissonance effect; that is, an upward evaluation of the chosen alternative and a downward evaluation of the rejected alternative, relative to no-choice conditions. Moreover, the relative upward evaluation of the chosen alternative increased as the expected delay decreased. Valence was not found to interact with other variables. The results are discussed in terms of differential salience of choice alternatives.     

Acceptance and Elimination Procedures in Choice: Noncomplementarity and the Role of Implied Status quo

The present research contrasts two seemingly complementary decision strategies: acceptance and elimination. In acceptance, a choice set is created by including suitable alternatives from an initial set of alternatives, whereas in elimination it is created by removing inappropriate alternatives from that same initial set. The research used realistic career decision-making scenarios and presented to respondents sets of alternatives that varied in their preexperimental strength values. Whereas complementarity of acceptance and elimination is implied by three standard (normative) assumptions of decision theory, we find a systematic discrepancy between the outcomes of these procedures: choice sets were larger in elimination than in acceptance. This acceptance–elimination discrepancy is directly tied to subcomplementarity. The central tenet of the theoretical framework developed here is that acceptance and elimination procedures imply different types of status quo for the alternatives, thereby invoking a different selection criterion for each procedure. A central prediction of the dual-criterion framework is that middling alternatives should be most susceptible to the type of procedure used. The present studies focus on this prediction which is substantiated by the results showing that middling alternatives yield the greatest discrepancy between acceptance and elimination. The implications of this model and findings for various research domains are discussed.     

Response Mode Bias in the Formation of Preference: Boundary Conditions of the Prominence Effect

Prior research has discovered that the most prominent attribute has greater influence on the formation of preference in choice versus matching tasks. We extend the research on this phenomenon, which is known as the prominence effect (Tversky, Sattath, & Slovic, 1988), by examining its generalizability and by providing insight into the psychological processes involved. The effects of task characteristics (i.e., the number of attributes and alternatives) and the effects of subject characteristics (i.e., processing goal) on the prominence effect were examined. In the first experiment we found that the prominence effect reverses when the number of attributes increases from two to four. That is, the prominent attribute is given greater weight in matching tasks rather than in choice tasks. A second experiment demonstrated that processing goal does influence the robustness of the prominence effect. We found that the influence of the prominent attribute on the formation of preference did not differ in choice and matching tasks when subjects′ processing goal was to form an overall impression of each of the alternatives. A third experiment, which explored the interaction between the response mode and processing goal in the four attribute case replicated this finding. Findings from Experiment 1 were also extended so that the reverse prominence effect was found when subjects processing goal was to memorize the information. Implications of these results are discussed.     

An Optimized Design of Choice Experiments: A New Approach for Studying Decision Behavior in Choice Task Experiments

In this paper, we present a new approach for the optimal experimental design problem of generating diagnostic choice tasks, where the respondent's decision strategy can be unambiguously deduced from the observed choice. In this new approach, we applied a genetic algorithm that creates a one‐to‐one correspondence between a set of predefined decision strategies and the alternatives of the choice task; it also manipulates the characteristics of the choice tasks. In addition, this new approach takes into account the measurement errors that can occur when the preferences of the decision makers are being measured. The proposed genetic algorithm is capable of generating diagnostic choice tasks even when the search space of possible choice tasks is very large. As proof‐of‐concept, we used this novel approach to generate respondent‐specific choice tasks with either low or high context‐based complexity that we operationalize by the similarity of alternatives and the conflict between alternatives. We find in an experiment that an increase in the similarity of the alternatives and an increase in the number of conflicts within the choice task lead to an increased use of non‐compensatory strategies and a decreased use of compensatory decision strategies. In contrast, the size of the choice tasks, measured by the number of attributes and alternatives, only weakly influences the strategy selection. Copyright © 2014 John Wiley & Sons, Ltd.     

Individual differences in the use of simplification strategies in a complex decision-making task

Decision time results were used to assess the strategies that 90 college undergraduates used in a complex decision-making task. Trend analyses revealed that the functions relating choice time to the number of choice alternatives in a set and the number of attributes comprising those alternatives contained linear (increasing) components. In addition, for a portion of the subjects, there was a quadratic effect of the number of attributes available to the decision maker on choice time, suggesting that these subjects adopted simplification strategies at high levels of task complexity. Reliable individual differences in these trend components were observed, consistent with individual differences in motivation and/or processing capacities. These individual differences were included in an information-processing model of decision behavior that captured the choice time data observed in this study. Subjects' ratings of apartments were used as a basis to assess the extent to which the use of simplification strategies resulted in preference reversals. Contrary to expectation, subjects whose choice times contained quadratic components demonstrated fewer preference reversals at high levels of information load.     

跨期决策是基于选项还是基于维度?

跨期决策有许多不同的理论模型,按其所指信息加工过程,可以分成基于选项的和基于维度的两类决策模型。前者认为各跨期选项被独立赋予一个主观价值,主观价值最大的选项被选择;而后者则认为选项是基于维度被比较的,比较中占优势的选项被选择。本文从行为结果、过程数据和对基于选项模型基本假设的质疑等角度综述了近来支持基于维度模型的证据;并提出未来需要进一步应用认知行为实验和眼动追踪技术对两类模型进行过程研究、在基于维度模型的指导下进行f MRI研究,以及加强损失领域的跨期决策研究以完善跨期决策模型。     

Multialternative decision field theory: a dynamic connectionist 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.     

The differential susceptibility of topographic map interpretation to influence from training

Performances on standardized map to landsurface matching tasks were compared following one of three training conditions. Subjects were explicitly instructed to represent the mapped landsurfaces as mental imagery arrays of selected features, and they were presented with sequences of landsurface drawings described as modelling image manipulation processes. Alternatively, subjects were presented with these same sequences, but without explicit instructions that they should be taken as modelling mental processes, or that any particular representational form should be used. Finally, control subjects were presented with neither explicit instructions on processing and representational form nor the sequences of landsurface drawings. Performance differences across the three groups were interpreted in terms of the use of mental imagery being strongly task-determined in such map matching tasks, but the choice of processing strategy being more susceptible to influence from training.     

Adaptive reasoning: Componential and eye movement analysis of geometric analogy performance

The present study explored individual differences in performance of a geometric analogies task. Whereas past studies employed true/false or two-alternative items, the present research included four-alternative items and studied eye movements and confidence judgements for each item performance as well as latency and error. Item difficulty proved to be a function of an interaction between the number of response alternatives and the number of elements in items, especially for subjects lower in fluid-analytic reasoning ability. Results were interpreted using two hypothesized performance strategies: constructive matching and response elimination. The less efficient of these, response elimination, seemed to be used more by lower ability subjects on more difficult items. While two previous theories resemble one or the other of these strategies, neither alone seems to capture the complexity of adaptive problem solving. It appears that a comprehensive theory should incorporate strategy shifting as a function of item difficulty and subject ability.Componential models, based in part on past research, revealed that a justification component was activated and deactivated depending upon the nature of the analogy being solved. In addition, two new components, spatial inference and spatial application, were identified as important on some items, suggesting that different geometric analogy items invoke different cognitive processing components. Thus, a comprehensive theory should also describe component activation and deactivation.     

Division of attention: The single-channel hypothesis revisited

Two experiments are reported in which subjects performed two forced-paced serial reaction time tasks separately and together at their maximum sustainable rates of information processing. Experiment 1 investigated the effects on the relationship between single- and dual-task performance of using tasks with the same or different input and output modality characteristics; an additional condition tested the effects on this relationship of using tasks with higher S-R compatibility. Experiment 2 investigated the effects on the relationship between single- and dual-task performance of varying information load (number of S-R alternatives). No significant differences were found in subjects' capacities to process information in single- and dual-task conditions. This finding was unaffected by: (a) the absolute information levels of the tasks, (b) whether inputs and/or outputs involved the same or different modalities, or (c) the level of S-R compatibility. The data from both experiments provide strong support for the single-channel hypothesis.     

Polarity correspondence: A general principle for performance of speeded binary classification tasks

Differences in performance with various stimulus-response mappings are among the most prevalent findings for binary choice reaction tasks. The authors show that perceptual or conceptual similarity is not necessary to obtain mapping effects; a type of structural similarity is sufficient. Specifically, stimulus and response alternatives are coded as positive and negative polarity along several dimensions, and polarity correspondence is sufficient to produce mapping effects. The authors make the case for this polarity correspondence principle using the literature on word-picture verification and then provide evidence that polarity correspondence is a determinant of mapping effects in orthogonal stimulus-response compatibility, numerical judgment, and implicit association tasks. The authors conclude by discussing implications of this principle for interpretation of results from binary choice tasks and future model development.     

Children’s choice strategies: the effects of age and task demands

Two experiments examined the effect of age and cognitive demands on children’s choice strategies. Children aged 8–9 and 12–13 years were asked to choose among either two or four products that differed in several attributes of varying importance to them. Choice tasks were designed to differentiate between the lexicographic and the equal-weighting strategies that the children used. The results showed that older children use the two strategies more appropriately than younger children do. The choice of strategies reflects the demands of the task. All children were more likely to use the strategies correctly when choosing between two alternatives and tended to rely on the relatively less cognitive-demanding lexicographic strategy to choose among four alternatives. However, they were less likely to employ the lexicographic strategy in simpler tasks and when its use was difficult to justify, indicating an adaptive use.     

The dimensional effects of missing information on choice processing

This paper demonstrates that choice processing may be different in missing information situations than in full information situations depending on whether inferences are used to fill in missing values and the overlap of the missing information itself. It is shown that when individuals do not form inferences to fill in missing values, fewer full attribute-based processes and more processes which accommodate for missing attribute values, alternative-based or given-dimension attribute-based, are used. It is also shown that when a processing shift due to missing information does occur, the overlap of the missing values will affect the type of shift that takes place. If overlap is high, a shift to given-dimension attribute-based processing is more likely, and when overlap is low, a shift to alternative-based processing is more likely. When individuals do form inferences to fill in missing values, processing is more similar to that in full information situations. Finally, it is shown that individuals will often partially fill in missing information, thus moderating the proposed effects.     

On adaptation, maximization, and reinforcement learning among cognitive strategies

Analysis of binary choice behavior in iterated tasks with immediate feedback reveals robust deviations from maximization that can be described as indications of 3 effects: (a) a payoff variability effect, in which high payoff variability seems to move choice behavior toward random choice; (b) underweighting of rare events, in which alternatives that yield the best payoffs most of the time are attractive even when they are associated with a lower expected return; and (c) loss aversion, in which alternatives that minimize the probability of losses can be more attractive than those that maximize expected payoffs. The results are closer to probability matching than to maximization. Best approximation is provided with a model of reinforcement learning among cognitive strategies (RELACS). This model captures the 3 deviations, the learning curves, and the effect of information on uncertainty avoidance. It outperforms other models in fitting the data and in predicting behavior in other experiments.