ALCOVE (attention learning covering map) is a connectionist model of category that incorporates an exemplar-based representation (Medin & Schaffer, 1978; Nosofsky, 1986) with error-driven (Gluck Bower, 1988; Rumelhart, Hinton, Williams, 1986). Alcove selectively attends to relevant stimulus dimensions, sensitive correlated can account for form base-rate neglect, does not suffer catastrophic forgetting, and exhibit 3-stage (U-shaped) high-frequency exceptions rules, whereas such effects are...

Bayesian estimation for 2 groups provides complete distributions of credible values the effect size, group means and their difference, standard deviations normality data. The method handles outliers. decision rule can accept null value (unlike traditional t tests) when certainty in estimate is high model comparison using Bayes factors). also yields precise estimates statistical power various research goals. software programs are free run on Macintosh, Windows, Linux platforms.

Psychological theories of categorization generally focus on either rule- or exemplar-based explanations. We present 2 experiments that show evidence both rule induction and exemplar encoding as well a connectionist model, ATRIUM, specifies mechanism for combining representation. In participants learned to classify items, most which followed simple rule, although there were few frequently occurring exceptions. Experiment 1 examined how people extrapolate beyond the range training. effect...

This article explains a decision rule that uses Bayesian posterior distributions as the basis for accepting or rejecting null values of parameters. focuses on range plausible indicated by highest density interval distribution and relation between this region practical equivalence (ROPE) around value. The also discusses considerations setting limits ROPE emphasizes analogous apply to thresholds p Bayes factors.

The use of Bayesian methods for data analysis is creating a revolution in fields ranging from genetics to marketing. Yet, results our literature review, including more than 10,000 articles published 15 journals January 2001 and December 2010, indicate that approaches are essentially absent the organizational sciences. Our article introduces science researchers describes why how they should be used. We multiple linear regression as framework offer step-by-step demonstration, software,...

Bayesian methods have garnered huge interest in cognitive science as an approach to models of cognition and perception. On the other hand, for data analysis not yet made much headway against institutionalized inertia 20th century null hypothesis significance testing (NHST). Ironically, specific perception may long endure ravages empirical verification, but generic will eventually dominate. It is time that became norm science. This article reviews a fatal flaw NHST introduces reader some...

A new connectionist model (named RASHNL) accounts for many "irrational" phenomena found in nonmetric multiple-cue probability learning, wherein people learn to utilize a number of discrete-valued cues that are partially valid indicators categorical outcomes. Phenomena accounted include cue competition, effects salience, utilization configural information, decreased learning when information is introduced after delay, and base rates. Experiments 1 2 replicate previous experiments on...

Adaptive network and exemplar-similarity models were compared on their ability to predict category learning transfer data. An exemplar-based (Kruschke, 1990a, 1990b, 1992) that combines key aspects of both modeling approaches was also tested. The incorporates an representation in which exemplars become associated categories through the same error-driven, interactive rules are assumed standard adaptive networks. Experiment 1, partially replicated extended probabilistic classification paradigm...

Substance dependent individuals (SDI) often exhibit decision-making deficits; however, it remains unclear whether the nature of underlying processes is same in users different classes drugs and these deficits persist after discontinuation drug use. We used computational modeling to address questions a unique sample relatively "pure" amphetamine-dependent (N = 38) heroin-dependent 43) who were currently protracted abstinence, 48 healthy controls (HC). A Bayesian model comparison technique,...

Abstract Probabilistic models based on Bayes' rule are an increasingly popular approach to understanding human cognition. Bayesian allow immense representational latitude and complexity. Because they use normative mathematics process those representations, define optimal performance a given task. This article focuses key mechanisms of information processing, provides numerous examples illustrating approaches the study We start by providing overview modeling networks. then describe three...

Abstract We compare how humans retell stories to ChatGPT retells in chains of three retellings by different people or accounts on ChatGPT. provides competent summaries the original narrative texts one step retelling. In subsequent few additional changes occur. Human retellers, contrast, reduce text incrementally and creating 55–60% novel words concepts (synsets) at each iteration. The both show very stable emotion ratings, which is a puzzle for human retellers given high degree inventions...