Choice Modeling

The Generalized Context Model (GCM; Nosofsky, 1986) is a highly successful theoretical model of perceptual categorization. The GCM proposes that stimuli are categorized based on their summed similarity to category exemplars stored in memory. In the GCM, similarity is a deterministic, decreasing, and non-linear function of distance in psychological space. However, the GCM uses a stochastic choice function – Luce’s Choice Rule (Luce, 1959) – to account for the variability in classification choices. While in general the GCM has excellent estimation properties, we argue that the free parameter in the stochastic choice function is under-determined both mathematically and conceptually. Using recent advances in probabilistic modeling, we explore the dependencies between parameters in the GCM, with a goal towards clarifying the role of response stochasticity. We also report the results of an empirical study of response stochasticity within perceptual categorization. This work aims to make stochasticity in categorization to become the object of substantive study and to further understand the mechanisms of the GCM.

This is an in-person presentation on July 19, 2026 (10:40 ~ 11:00 EDT).

The 2N-ary Choice Tree (2NCT) model (Wollschlaeger and Diederich, 2012, 2020) and the Cube model (Mallahi-Karai and Diederich, 2019, 2023) are dynamic-stochastic approaches for decision making situations with multiple alternatives. The 2NCT model is a dynamic stochastic model formalized as a random walk on a tree. It shares several features of other stochastic models on decision making such as including initial biases for any of the choice alternatives, updating preferences over time or initiation of a response when a decision criterion is met . Its distinct assumption is that it establishes two counters for each alternative for tracking evidences in favor of choosing a specific alternative and a negative counter for tracking evidence against choosing that alternative. It provides a mechanism to predict a preference order of the the offered alternative The multi-episode Cube model postulates that best–worst choice task is the outcome of sequential choices made in a number of episodes allowing the alternatives to be ranked from best to worst or from worst to best. The underlying model is a multivariate Wiener process with drift issued from a point in the unit cube, where episodes are defined in terms of a sequence of stopping times. Both models make predictions with respect to choice probabilities and (mean) choice response times. It is shown how the models can be implemented using Markov chains and how they are tested on data.

This is an in-person presentation on July 19, 2026 (11:00 ~ 11:20 EDT).

Cognitive models provide profound insights into the psychological processes underlying behavior. However, a significant limitation has constrained their application: the reliance on extensive, repeated-trial data from each participant. The data-hungry nature of cognitive models has largely precluded their application to infrequent but consequential one-shot decisions common in economic, social, and clinical contexts, and has excluded populations unable to complete lengthy experiments. Here, we address this methodological constraint by proposing a conceptual shift: instead of requiring many trials from a few individuals, we leverage few trials from many individuals. By treating between-subject variability as a source of information, we demonstrate that cognitive models can be successfully fit to one-shot data. Through a series of simulations, we first establish that we can recover known parameter values from single-trial data. We then validate the approach empirically by showing that with only a single trial per participant it can replicate canonical findings; namely, the speed-accuracy trade-off, the influence of food quality and expectations on choice, and the task-specificity of linear vs. non-linear numeric representations. This work overcomes a limitation to the widespread application of cognitive models and opens new frontiers for understanding the cognitive mechanisms underlying real-world choices.

This is an in-person presentation on July 19, 2026 (11:20 ~ 11:40 EDT).

A previous publication illustrated how to measure a thought using DMSD. Observers were presented with stimuli to choose between. Then they also illustrated the difference between the two stimuli by extending a horizontal line centered on their display screen. The average distances between two stimuli are perfectly correlated with the parameters in equations predicting choice probabilities. This shows that the probability of choice can be determined from the directly measured stimulus differences. The equation determining choice probability is due created by the characterization of the mental process of comparison that leads to a choice. The choice process parameters turn out to be part of the equation for the choice probabilities. This aspect of choice probability is largely overlooked because the process leading to a response is unknown to be a contributor to the choice probability. To understand cause and effect, an analysis of data must reveal the parameters of the decision process itself separate from unknown parameters that are the sought for measures of internal stimuli.

This is an in-person presentation on July 19, 2026 (11:40 ~ 12:00 EDT).