Exemplar-based judgment or direct recall: On a problematic procedure for estimating parameters in exemplar models of quantitative judgment

Abstract

Exemplar models are often used in research on multiple-cue judgments to describe the underlying process of participants’ responses. In these experiments, participants are repeatedly presented with the same exemplars (e.g., poisonous bugs) and instructed to memorize these exemplars and their corresponding criterion values (e.g., the toxicity of a bug). We propose that there are two possible outcomes when participants judge one of the already learned exemplars in some later block of the experiment. They either have memorized the exemplar and their respective criterion value and are thus able to recall the exact value, or they have not learned the exemplar and thus have to judge its criterion value, as if it was a new stimulus. We argue that psychologically, the judgments of participants in a multiple-cue judgment experiment are a mixture of these two qualitatively distinct cognitive processes: judgment and recall. However, the cognitive modeling procedure usually applied does not make any distinction between these processes and the data generated by them. We investigated potential effects of disregarding the distinction between these two processes on the parameter recovery and the model fit of one exemplar model. We present results of a simulation as well as the reanalysis of five experimental data sets showing that the current combination of experimental design and modeling procedure can bias parameter estimates, impair their validity, and negatively affect the fit and predictive performance of the model. We also present a latent-mixture extension of the original model as a possible solution to these issues.

Measuring the mixture of rule-based and exemplar-based processes in judgment: A hierarchical Bayesian approach.

“…While the state-of-the-art Bayesian hierarchical approach improves upon problems of parameter estimation of the original RulEx-J model as a measurement model, the Bayesian framework used in this article also offers new possibilities to implement and then compare different model variants to answer theoretical questions. For instance, by incorporating a learning process (e.g., Hoffmann et al, 2019), adding possible contamination processes (e.g., Zeigenfuse & Lee, 2010), more complex rule-process or exemplar-process models (e.g., Izydorczyk & Bröder, 2021), integrating additional sources of information or covariates (e.g., mouse tracking, eye tracking, electroencephalogram [EEG]).…”

“…In the first simulation, the α parameter values were drawn from a uniform β (1, 1) distribution and in the second simulation from a uniform β (1, 1), a β (5, 5), or peaked β (15, 15) distribution, simulating different levels of underlying similarities between participants (see Figure 1, for an illustration of the resulting distributions). The s parameter values were drawn from a slightly skewed β (3, 5) distribution which reflects a sensible range of s parameter values found in experimental studies (Izydorczyk & Bröder, 2021). The parameter values for the cue weights ω j were drawn from a truncated normal distribution with μ = ωjgen, σ = 1, an upper bound of 100, and a lower bound of 1.…”

Eye movements as a tool to investigate exemplar retrieval in judgments

“…According to the 'division of labor' hypothesis (i.e., Juslin et al, 2008), people either rely on an exemplar-based strategy or a rule-based strategy. However, further research proposes that judgments may be formed by a 'blending' or mixture of both exemplar-based and rule-based processes within a single judgment (Albrecht et al, 2020;Bröder et al, 2017;Erickson & Kruschke, 1998;Herzog & von Helversen, 2018;Hoffmann et al, 2014;Izydorczyk & Bröder, 2021;Schlegelmilch et al, 2022;von Helversen et al, 2014). Thereafter, people can adaptively exploit both their memories of previous experiences and combine information following a rule-like integration process.…”

“…In the second step, the retrieved judgment is adjusted based on abstracted cue knowledge. Blending models like the measurement model RuleEx-J for judgments (Bröder et al, 2017;Izydorczyk & Bröder, 2021) assume that an exemplar-and a rule-based judgment process act in parallel. The resulting judgment is a weighted combination of both interim judgments.…”

What is the airspeed velocity of an unladen swallow? Modeling numerical judgments of realisitc stimuli

“…Two of the most prominent cognitive process models describing multiple-cue judgment processes are rule and exemplar models. In exemplar models, the judgment of a given object is generated by retrieving similar objects from long-term memory and by forming a similarity-weighted average of their criterion values (e.g., Izydorczyk & Bröder, 2021;Juslin et al, 2003;Medin & Schaffer, 1978;Nosofsky, 1984). In rule-based models, the judgments rely on abstract knowledge about cue-criterion relations, such as a linear combination of cues (e.g., Brehmer, 1994;Einhorn et al, 1979;Juslin et al, 2003).…”