William M. Goldstein

Content and discontent: Indications and implications of domain specificity in preferential decision making.

Publisher Summary This chapter explores that research on preferential decision making has been and largely continues to be dominated by the use of a particular kind of stimulus material: simple monetary gambles. There is a critical presentation of the metatheory that guides most research on preferential judgment and choice, which is called “the gambling metaphor,” but to do so specifically with respect to the issue of content effects. Others have voiced concerns at the level of general metatheory, but these have related mainly to other issues. The chapter reviews several areas of psychology in which the semantic content of stimuli has been found to be an important determinant of behavior: in memory, animal learning, categories and concepts, deductive reasoning, problem solving and expertise, and cognitive development. The psychological mechanisms underlying preferential decision making overlap sufficiently with those in other areas of psychology that decision researchers should draw from the substantive results concerning the effects of semantic content. It also reviews the research on preferential decision making that bears on the issue of content dependence. There is also some discussion about the views of what a theory of content effects in preferential decision making, which might look like if it is to maintain a reasonable degree of parsimony.

Judgments of relative importance in decision making : global vs local interpretations of subjective weight

Abstract Because most choices involve alternatives that have both advantages and disadvantages, many issues in decision research concern the way people evaluate trade-offs among conflicting goals or attributes. The attributes of the alternatives are often perceived to be of unequal importance to the decision, leading to questions about peoples understanding and use of “relative importance” information to characterize their own and other peoples decisions. Global interpretations of relative importance consider it to be a fixed attitude of the decision maker, a stable characteristic that does not depend on the particular stimuli involved, provided that the stimuli do not disturb the persons implicit contextual assumptions. By contrast, local interpretations of relative importance hold it to be an assessment, like preference itself, that depends on both the personal characteristics of the decision maker and the stimuli. These interpretations are compared in an experiment that asked subjects to (1) provide their own relative importance ratings and preference orders, (2) use the relative importance ratings of a target person to infer the target persons preference order, and (3) infer the relative importance ratings of a target person based on the target persons preference order. The way that peoples decision patterns covaried with their relative importance judgments across stimulus sets largely violated global interpretations, although some of the evidence was ambiguous. Specific interpretations of relative importance are discussed, as well as factors that seem likely to affect peoples selection of an interpretation.

Stochastic PERT networks: OP diagrams, critical paths and the project completion time

Abstract Fisher and Goldstein [5–7] have recently derived closed form expressions for the expectation, variance and distribution of the project completion time of stochastic PERT networks when the durations of the tasks in the network can be written as mutually independent, probability mixtures of general-gamma [8] random variables (a more general result is described in [9]). In this article, it is shown that their methods can also be used to obtain the probability that a particular path in the PERT network is critical [6]. One possible criticism of their approach is that computation of both the probability that a particular path is critical and the summary characteristics of the reaction time distribution require the translation of a PERT network into an OP (Order-of-ftocessing) diagram. Currently, this translation can only be done manually, a very laborious process for even moderately complex networks. In this article, an algorithm is developed which can be used to generate the OP diagram for a given PERT network. The algorithm has been implemented in FORTRAN. Once the OP diagram is constructed, computation of both the probability that a particular path is critical and the expectation, variance and distribution of the project completion time requires only simple matrix operations when the durations of the tasks in the network are mutually independent probability mixtures of general-gamma random variables. More complex calculations are required when other distributions of the task durations are involved.

And let us not Forget Memory: The Role of Memory Processes and Techniques in the Study of Judgment and Choice

Publisher Summary This chapter considers the two main metatheoretical frameworks of J/DM research that is, strategy selection and the perceptual framework, for which there is some discussion about the potential contributions of memory research. With respect to strategy selection, the chapter demonstrates recent models of memory representation and memory processes that have the potential to enlarge the set of explanatory constructs that J/DM researchers have available to account for information integration. Although information integration is usually thought to be accomplished by an effortful decision strategy, the formation of overall impressions is the relatively effortless by-product of memory representations and storage operations. The chapter describes the theoretical and methodological implications of distinctions based on the content of memory. It also explores several directions for theory development that integrate memory processes and judgment or decision processes in different ways and at different levels of analysis. Researchers in other areas of cognitive psychology routinely acknowledge interconnections among different cognitive processes.

Linking together different measures of preference: A dynamic model of matching derived from decision field theory

Abstract Two frequently used tasks for measuring preferences among gambles are choice and selling price tasks. However, the rank orders observed with these tasks do not agree, and these disagreements are called preference reversals. In this article, we propose an extension of decision field theory that was originally designed to account for choice probability and the distribution of choice response times. In this extension, we show how this same theory can be used to derive predictions for the distribution of values produced by selling price tasks. This model not only accounts for the basic preference reversal results but also can explain the effects of various information processing factors on preference reversals including time, effort, and practice. We conclude by summarizing the advantages of the decision field matching model over two earlier models of choice and selling prices—expression theory and the contingent weighting model.

Decomposable threshold models

Abstract Subjects who must decide whether an alternative is acceptable, whether a stimulus belongs to a specified category, or whether a stimulus is recognized as previously displayed, have often been modeled as attempting to determine whether a multidimensional stimulus exceeds a threshold. This paper examines and axiomatizes some general models for these discrete judgment tasks. The threshold models considered, viewed as models of preference, generalize the conjunctive and disjunctive rules, among other noncompensatory and compensatory decision rules. Viewed as models of categorization, the threshold models generalize independent cue models and the notion of linear separability of categories. For each threshold model considered, necessary and sufficient conditions are presented. The axiomatic analyses direct attention to the features of the models that can be subjected to empirical test. It turns out that the most general form considered is too general to be falsifiable. In each restricted case, the key feature of the model turns out to be a variant of a “context-independence” condition that is highly fasifiable.

Stochastic networks as models of cognition: derivation of response time distributions using the order-of-processing method

Abstract Researchers in cognitive psychology have devoted a great deal of attention to the temporal nature of cognitive processes. Hypotheses are tested by comparing obtained and theoretical response time distributions for various experimental conditions. This paper extends and refines a method for deriving theoretically expected response time distributions, so that it can be applied under weaker assumptions and to a wider class of mental architectures. Specifically, the Order-of-Processing (OP) method, originally developed to analyze stochastic PERT networks, is extended to apply to much more general structures. The method is also extended to handle more general distributions of cognitive process durations than previously considered. As an example of the OP method, OP diagrams are constructed for models which are not representable as stochastic PERT networks—limited-capacity and self-terminating models of visual search.

Stochastic networks as models of cognition: deriving predictions for resource-constrained mental processing

Abstract One approach to the analysis of task performance is to focus on the identity and arrangement of subtasks. For our purposes, we consider a theory of task performance to consist of: (1) a decomposition of the task into a set of hypothesized subtasks or mental processes; (2) hypothesized structural constraints concerning the order of execution of mental processes (i.e., which of the mental processes must occur sequentially and which can occur concurrently), whether subtasks can be interrupted and resumed, whether potentially concurrent subtasks are prioritized, etc.; and (3) temporal hypotheses concerning the time required to complete each of the subtasks. A large class of such theories can be represented by Order-of-Processing (OP) diagrams. Previous research has indicated how to compute the theoretically expected response-time distribution for OP-representable tasks, provided that processing resources are allocated in such a manner that each subtask undergoing execution progresses at a constant, unit rate. Here, we present results for OP-representable tasks in which processing resources (e.g., attention) can be differentially and dynamically allocated to the execution of mental processes, the exact allocation of resources determining the rate at which each subtask is executed at any given moment.

Expression theory and the measurement of apparently labile values

Systematic inconsistencies in judgments and choices have been attributed to either the encoding of information and/or its evaluation. However, anomalics in decision making can also result from the way people express their internal experiences on different overt response scales. A model that relates underlying preferences for gambles to a variety of observable judgments and choices (e.g. binary choices, judgments of minimum selling price, ratings of attractiveness) is developed. The model, called Expression Theory, explains how stable and consistent underlying preferences can result in inconsistent overt responses, including “preference reversal” phenomena. Thus, one can choose alternative A over B, but rate B as more attractive than A. An axiomatic foundation is developed for Expression Theory that gives insight into the structure of the model and suggests further empirical work. The relative importance of the axioms to the substantive theory is discussed, along with issues relevant to the measurement of the numerical scales.