Exemplar Model Research Articles

Is categorization performance really intact in amnesia? A meta-analysis

Most published studies of category learning in amnesia have reported intact categorization performance. These results have been used to challenge single-system accounts of categorization and recognition, in which a single representational system mediates performance in these two tasks. Many of the published studies, however, have shown a numerical advantage for controls over amnesics and often have had low statistical power. A meta-analysis was conducted to assess whether this numerical advantage is significant when the data are pooled across studies. This analysis indicates that amnesic subjects do, in fact, show deficits in categorization tasks, which is consistent with single-system exemplar model predictions.

False prototype enhancement effects in dot pattern categorization.

Results from the classic dot pattern distortion paradigm have sometimes yielded prototype enhancement effects that could not be accounted for by exemplar models of categorization. However, in these experiments the status of the prototype was confounded with certain stimulus-specific properties as well as with the frequency of presentation of the prototype during testing. In two mock-subliminal experiments, participants made categorization judgments to patterns that were generated as prototypes, low-level distortions, or high-level distortions. The participants rated the prototypes as being more likely to be members of a category, although no patterns were presented during training, and there was no objective category structure. In two other experiments, greater prototype enhancement effects were observed when the prototype and low-level distortions were presented with greater frequency during transfer. These results suggest that classic prototype enhancement effects may not be due to the abstraction of a prototype at time of original learning, but rather to other factors not formalized in extant models.

Sequence effects in the categorization of tones varying in frequency.

In contrast to exemplar and decision-bound categorization models, the memory and contrast models described here do not assume that long-term representations of stimulus magnitudes are available. Instead, stimuli are assumed to be categorized using only their differences from a few recent stimuli. To test this alternative, the authors examined sequential effects in a binary categorization of 10 tones varying in frequency. Stimuli up to 2 trials back in the sequence had a significant effect on the response to the current stimulus. The effects of previous stimuli interacted with one another. A memory and contrast model, according to which only ordinal information about the differences between the current stimulus and recent preceding stimuli is used, best accounted for these data.

SPICE for SPACE trials, risk analysis, and process improvement

AbstractAs part of a programme for software process improvement sponsored by the European Space Agency (ESA), an ISO 15504 (1998) conformant method for software process assessment has been developed. [Work supported under ESTEC Contract Number 10 662/93/NL/NB CCN5.] Called SPICE for SPACE, or S4S (Völcker and Cass 2002, Cass et al. 2000), the method includes an assessment model based on the ISO/IEC TR 15504 exemplar model with space software practices incorporated.To promote SPICE for SPACE, ESA has sponsored a series of trial assessments using the S4S method. Since February 2000, eight space software suppliers have been assessed in Austria, Belgium, Denmark, Germany, Great Britain, Italy, and Sweden.The first part of this paper describes the trial assessment programme to date, summarising the assessments that have taken place, the assessment results, the improvement efforts initiated in the host organisations, and the feedback gathered to improve the S4S method.The second part of this paper presents a new approach for software process improvement based on a risk analysis of assessment results. Developed within the same initiative, this method sets the basis for integrating process‐oriented risk analysis into the S4S method, where, along with the process and capability dimensions, risk can be considered as a new third dimension. The outcomes of the risk analysis lead to the selection and prioritisation of pre‐defined generic improvement actions. Alternatively, the risk analysis presented here can be applied in the context of supplier evaluation or risk estimation for investors in software organisations. Copyright © 2004 John Wiley & Sons, Ltd.

Situated simulation in the human conceptual system

Four theories of the human conceptual system—semantic memory, exemplar models, feed‐forward connectionist nets, and situated simulation theory—are characterised and contrasted on five dimensions: (1) architecture (modular vs. non‐modular), (2) representation (amodal vs. modal), (3) abstraction (decontextualised vs. situated), (4) stability (stable vs. dynamical), and (5) organisation (taxonomic vs. action–environment interface). Empirical evidence is then reviewed for the situated simulation theory, and the following conclusions are reached. Because the conceptual system shares mechanisms with perception and action, it is non-modular. As a result, conceptual representations are multi-modal simulations distributed across modality‐specific systems. A given simulation for a concept is situated, preparing an agent for situated action with a particular instance, in a particular setting. Because a concept delivers diverse simulations that prepare agents for action in many different situations, it is dynamical. Because the conceptual system’s primary purpose is to support situated action, it becomes organised around the action–environment interface.

Learning foreign vowels.

Two hypotheses have recently been put forward to account for listeners' ability to distinguish and learn contrasts between speech sounds in foreign languages. First, Best's Perceptual Assimilation Model and Flege's Speech Learning Model both predict that the ease with which a listener can tell one non-native phoneme from another varies directly with the extent to which these sounds assimilate to different native phonemes (Best, 1994; also Best, McRoberts, & Goodell, 2001; Flege, 1991). Second, Logan, Lively, & Pisoni (1991) have argued that training listeners to identify non-native phonemes teaches them sets of exemplars rather than more abstract distinctive feature values. I report here the results of three sets of experiments designed to test these hypotheses, in which American English listeners were trained to categorize German nonlow vowels. The first set of experiments show that some instances of the same contrast between German vowels are more easily discriminated than others, a result incompatible with the predictions of either Best's or Flege's models, but compatible with the alternative category recognition interpretation. The second set of experiments reveals effects of contextual and speaker variation on listeners' ability to learn [tense] but not [high] contrasts between foreign vowels, and are thus at least partly compatible with an exemplar model of foreign category learning (Pisoni, Lively, & Logan, 1994; also Nosofsky, 1986). The third set of experiments compares the predictions of Nosofsky's (1986) selective attention exemplar model of category learning with those of a feature learning model in tests of listeners' learning the natural classes to which the German vowels belong. The results are mixed: listeners learned the features that define the natural classes of [+/- high] and [+/- back] vowels, but could have learned either the feature that defines the natural classes of [+/- tense] vowels or sets of [+/- tense] exemplars. Natural classes defined by abstract distinctive feature values are thus learnable, even if their membership is phonetically polymorphous.

Models of Thinking, Learning, and Teaching in Games

Noncooperative game theory combines strategic thinking, best-response, and mutual consistency of beliefs and choices (equilibrium). Hundreds of experiments show that in actual behavior these three forces are limited, even when subjects are highly motivated and analytically skilled (Camerer, 2003). The challenge is to create models that are as general, precise, and parsimonious as equilibrium, but which also use cognitive details to explain experimental evidence more accurately and to predict new regularities. This paper describes three exemplar models of behavior in one-shot games (thinking), learning over time, and how repeated “partner” matching affects behavior (teaching) (see Camerer et al., 2002b).

Effects of choice and relative frequency elicitation on overconfidence: further tests of an exemplar‐retrieval model

AbstractAn experiment is reported in which participants rendered judgments regarding the disease states of hypothetical patients. Participants either reported likelihoods that patients had the target disease (no choice), or classified patients into disease categories and then reported likelihoods that their classifications were correct (choice included). Also, participants' likelihood judgments were made in response to either a probability probe question, or a relative frequency probe. Two distinct exemplar‐memory models were compared on their ability to predict overconfidence under these procedures. Both propose that people learn and judge by storing and retrieving examples. The exemplar retrieval model (ERM) proposes that amount of retrieval drives choice inclusion and likelihood probe effects. The alternative model assumes that response error mediates choice inclusion effects. Choice inclusion and the relative frequency probe reduced overconfidence, but the combined effects were subadditive. Only the ERM predicted this pattern, and it further provided good quantitative fits to these results. Copyright © 2003 John Wiley & Sons, Ltd.

Double Dissociations, Models and Serial Position Curves'

What inferences do double dissociations permit? Here we suggest that in drawing such inferences it may be useful to utilise exemplar models of memory and categorisation that have been independently developed outside neuropsychology. We provide examples of the successful application of such models to dissociation data, and then address serial position effects in amnesia.

Fruits and vegetables categorized: an application of the generalized context model.

In the study reported in this paper, we investigated the categorization of well-known and novel food items in the categories fruits and vegetables. Predictions based on Nosofsky's (1984,1986) generalized context model (GCM), on a multiplicative-similarity prototype model, and on an instantiation model as applied in Storms, De Boeck, and Ruts (2001) were compared. Despite suggestions in the literature that prototype models predict categorization from large categories better than exemplar models do, our results showed that the exemplar-based GCM yielded clearly better predictions than did a (multiplicative-similarity) prototype model.

Are there representational shifts during category learning?

Early theories of categorization assumed that either rules, or prototypes, or exemplars were exclusively used to mentally represent categories of objects. More recently, hybrid theories of categorization have been proposed that variously combine these different forms of category representation. Our research addressed the question of whether there are representational shifts during category learning. We report a series of experiments that tracked how individual subjects generalized their acquired category knowledge to classifying new critical transfer items as a function of learning. Individual differences were observed in the generalization patterns exhibited by subjects, and those generalizations changed systematically with experience. Early in learning, subjects generalized on the basis of single diagnostic dimensions, consistent with the use of simple categorization rules. Later in learning, subjects generalized in a manner consistent with the use of similarity-based exemplar retrieval, attending to multiple stimulus dimensions. Theoretical modeling was used to formally corroborate these empirical observations by comparing fits of rule, prototype, and exemplar models to the observed categorization data. Although we provide strong evidence for shifts in the kind of information used to classify objects as a function of categorization experience, interpreting these results in terms of shifts in representational systems underlying perceptual categorization is a far thornier issue. We provide a discussion of the challenges of making claims about category representation, making reference to a wide body of literature suggesting different kinds of representational systems in perceptual categorization and related domains of human cognition.

PROBabilities from EXemplars (PROBEX): a “lazy” algorithm for probabilistic inference from generic knowledge

PROBEX (PROBabilities from EXemplars), a model of probabilistic inference and probability judgment based on generic knowledge is presented. Its properties are that: (a) it provides an exemplar model satisfying bounded rationality; (b) it is a “lazy” algorithm that presumes no pre-computed abstractions; (c) it implements a hybrid-representation, similarity-graded probability. We investigate the ecological rationality of PROBEX and find that it compares favorably with Take-The-Best and multiple regression (Gigerenzer, Todd, & the ABC Research Group, 1999). PROBEX is fitted to the point estimates, decisions, and probability assessments by human participants. The best fit is obtained for a version that weights frequency heavily and retrieves only two exemplars. It is proposed that PROBEX implements speed and frugality in a psychologically plausible way.

Word frequency and context of use in the lexical diffusion of phonetically conditioned sound change

The literature on frequency effects in lexical diffusion shows that even phonetically gradual changes that in some cases are destined to be lexically regular show lexical diffusion while they are in progress. Change that is both phonetically and lexically gradual presents a serious challenge to theories with phonemic underlying forms. An alternate exemplar model that can account for lexical variation in phonetic detail is outlined here. This model predicts that the frequency with which words are used in the contexts for change will affect how readily the word undergoes a change in progress. This prediction is tested on data from /t, d/ deletion in American English. Finally, the effect of bound morphemes on the diffusion of a sound change is examined. The data suggest that instances of a bound morpheme can affect the rate of change for that morpheme overall.

Exemplar theory's predicted typicality gradient can be tested and disconfirmed.

One of exemplar theory's central predictions concerns the shape of typicality gradients. The typicality gradient it predicts is a consequence of its exemplar-based comparisons and appears no matter how the theory is evaluated. However, this predicted typicality gradient does not fit the empirical typicality gradients obtained in an influential version of the dot-distortion category task, and this is true even when the exemplar model is made more flexible and mathematically powerful. Thus, exemplar theory is disconfirmed in this domain of categorization. In contrast, prototype theories are consistent with the empirically obtained typicality gradients.

PROBabilities from EXemplars (PROBEX): a “lazy” algorithm for probabilistic inference from generic knowledge

AbstractPROBEX (PROBabilities from EXemplars), a model of probabilistic inference and probability judgment based on generic knowledge is presented. Its properties are that: (a) it provides an exemplar model satisfying bounded rationality; (b) it is a “lazy” algorithm that presumes no pre‐computed abstractions; (c) it implements a hybrid‐representation, similarity‐graded probability. We investigate the ecological rationality of PROBEX and find that it compares favorably with Take‐The‐Best and multiple regression (Gigerenzer, Todd, & the ABC Research Group, 1999). PROBEX is fitted to the point estimates, decisions, and probability assessments by human participants. The best fit is obtained for a version that weights frequency heavily and retrieves only two exemplars. It is proposed that PROBEX implements speed and frugality in a psychologically plausible way.

Recognizing spatial patterns: a noisy exemplar approach

Models of categorization typically rely on the use of stimuli composed of well-defined dimensions (e.g., Ashby & Maddox (1998) in Choice, decision, and measurement: Essays in honor of R. Duncan Luce, p. 251–301, Mahwah, NJ: Erlbaum). We apply a similar approach to the analysis of recognition memory. Using a version of short-term recognition paradigm (Sternberg, Science 153 (1966) 652), we asked whether NEMO Sternberg's, a noisy exemplar summed-similarity model, could account for variation in mean performance on individual trials. NEMO provided a very good overall fit to recognition data from three experiments. However, its failure to fit data for certain lists of stimuli suggested a revision of the summed-similarity assumption. Our model-based analysis showed that subjects used interitem similarity, in addition to probe-item similarity, as the basis for their decisions. This represents a major departure from existing recognition models that assume subjects' judgments depend exclusively on the summed similarity of the probe to the study items.

Mixture Models of Categorization

Many currently popular models of categorization are either strictly parametric (e.g., prototype models, decision bound models) or strictly nonparametric (e.g., exemplar models) (F. G. Ashby & L. A. Alfonso-Reese, 1995, Journal of Mathematical Psychology, 39, 216–233). In this article, a family of semiparametric classifiers is investigated where categories are represented by a finite mixture distribution. The advantage of these mixture models of categorization is that they contain several parametric models and nonparametric models as a special case. Specifically, it is shown that both decision bound models (F. G. Ashby & W. T. Maddox, 1992, Journal of Experimental Psychology: Human Perception and Performance, 16, 598–612; 1993, Journal of Mathematical Psychology, 37, 372–400) and the generalized context model (R. M. Nosofsky, 1986, Journal of Experimental Psychology: General, 115, 39–57) can be interpreted as two extreme cases of a common mixture model. Furthermore, many other (semiparametric) models of categorization can be derived from the same generic mixture framework. In this article, several examples are discussed and a parameter estimation procedure for fitting these models is outlined. To illustrate the approach, several specific models are fitted to a data set collected by S. C. McKinley and R. M. Nosofsky (1995, Journal of Experimental Psychology: Human Perception and Performance, 21, 128–148). The results suggest that semi-parametric models are a promising alternative for future model development.

Rule-based extrapolation in perceptual categorization.

Erickson and Kruschke (1998) provided a demonstration that in certain situations people will classify novel stimuli according to an extrapolated rule, even when the most similar training exemplar is an exception to the rule. This result challenged exemplar models. Nosofsky and Johansen (2000) have called this finding into question by offering an exemplar-based explanation for those data based on the perceptual features of the stimuli. Here, we describe the results of a new experiment that yields results similar to those found previously without the questionable perceptual features: Participants who learn to classify all the training stimuli have patterns of generalization that indicate a combination of rule and exemplar representation. ATRIUM, a hybrid rule and exemplar model (Erickson & Kruschke, 1998), is shown to account for these data much better than ALCOVE, an exemplar model (Kruschke, 1992). Moreover, four alternate exemplar explanations, including one suggested by Nosofsky and Johansen, cannot account for our new findings.

Exemplar and prototype models revisited: Response strategies, selective attention, and stimulus generalization.

J. D. Smith and colleagues (J. P. Minda & J. D. Smith, 2001; J. D. Smith & J. P. Minda, 1998,2000; J. D. Smith, M. J. Murray, & J. P. Minda, 1997) presented evidence that they claimed challenged the predictions of exemplar models and that supported prototype models. In the authors' view, this evidence confounded the issue of the nature of the category representation with the type of response rule (probabilistic vs. deterministic) that was used. Also, their designs did not test whether the prototype models correctly predicted generalization performance. The present work demonstrates that an exemplar model that includes a response-scaling mechanism provides a natural account of all of Smith et al.'s experimental results. Furthermore, the exemplar model predicts classification performance better than the prototype models when novel transfer stimuli are included in the experimental designs.

Comparing prototype-based and exemplar-based accounts of category learning and attentional allocation.

Exemplar theory was motivated by research that often used D. L. Medin and M. M. Schaffer's (1978) 5/4 stimulus set. The exemplar model has seemed to fit categorization data from this stimulus set better than a prototype model can. Moreover, the exemplar model alone predicts a qualitative aspect of performance that participants sometimes show. In 2 experiments, the authors reexamined these findings. In both experiments, a prototype model fit participants' performance profiles better than an exemplar model did when comparable prototype and exemplar models were used. Moreover, even when participants showed the qualitative aspect of performance, the exemplar model explained it by making implausible assumptions about human attention and effort in categorization tasks. An independent assay of participants' attentional strategies suggested that the description the exemplar model offers in such cases is incorrect. A review of 30 uses of the 5/4 stimulus set in the literature reinforces this suggestion.