Learning concepts when instances never repeat.

Three experiments explored the learning of categories where the training instances either repeated in each training block or appeared only once during the entire learning phase, followed by a classification transfer (Experiment 1) or a recognition transfer test (Experiments 2 and 3). Subjects received training instances from either two (Experiment 2) or three categories (Experiments 1-3) for either 15 or 20 training blocks. The results showed substantial learning in each experiment, with the notable result that learning was not slowed in the non-repeating condition in any of the three experiments. Furthermore, subsequent transfer was marginally better in the non-repeating condition. The recognition results showed that subjects in the repeat condition had substantial memory for the training instances, whereas subjects in the non-repeat condition had no measurable memory for the training instances, as measured either by hit and false-alarm rates or by signal detectability measures. These outcomes are consistent with prototype models of category learning, at least when patterns never repeat in learning, and place severe constraints on exemplar views that posit transfer mechanisms to stored individual traces. A formal model, which incorporates changing similarity relationships during learning, was shown to explain the major results.

Feature-to-Feature Inference Under Conditions of Cue Restriction and Dimensional Correlation.

The present study explored feature-to-feature and label-to-feature inference in a category task for different category structures. In the correlated condition, each of the 4 dimensions comprising the category was positively correlated to each other and to the category label. In the uncorrelated condition, no correlation existed between the 4 dimensions comprising the category, although the dimension to category label correlation matched that of the correlated condition. After learning, participants made inference judgments of a missing feature, given 1, 2, or 3 feature cues; on half the trials, the category label was also included as a cue. The results showed superior inference of features following training on the correlated structure, with accurate inference when only a single feature was presented. In contrast, a single-feature cue resulted in chance levels of inference for the uncorrelated structure. Feature inference systematically improved with number of cues after training on the correlated structure. Surprisingly, a similar outcome was obtained for the uncorrelated structure, an outcome that must have reflected mediation via the category label. A descriptive model is briefly introduced to explain the results, with a suggestion that this paradigm might be profitably extended to hierarchical structures where the levels of feature-to-feature inference might vary with the depth of the hierarchy.

Identity categories and transformational paths for face changes across the age spectrum.

How faces change across lengthy time periods and whether the changing appearance of a face functions as an identity category was investigated in two experiments. In Experiment 1, the faces of 15 individuals were multidimensionally scaled at each of seven age epochs (roughly <6 months to 75 years of age) and correlated with the same persons, but at different ages. In Experiment 2, three individuals at each of seven age epochs were multidimensionally scaled, and analyses explored the conceptual structure and transformational path of each person within the space. The results revealed that even the earliest age epochs (<6 months) were correlated with the later epochs, except for the most extreme age, with the correlation highest at intermediate ages. Transformational paths were found that emerged and terminated at haphazard points in the space. Results are discussed in terms of principled change in categories whose members systematically change across their life cycle, including suggestions for the need to incorporate systematic change in current theories of categorization.

Isomorphic categories.

Learning and transfer were investigated for a categorical structure that could be mapped without loss of information from 1 sensory modality to another. The category space was composed of 3 nonoverlapping, linearly separable categories whose members were structured, temporal events. Each stimulus was composed of a sequence of on-off events that varied in duration and number of subevents (complexity). Categories were learned visually, haptically, or auditorily and transferred to the same or another modality. Despite the.isomorphism across modalities, significant differences appeared in both learning and transfer. The visual modality showed an early learning advantage, with information on the transfer test preserved best when encoded visually during learning, worst when encoded haptically, with auditory encoding intermediate. False recognition rates were elevated when categories were learned haptically and transferred to another modality. In general, classification accuracy was highest for the category prototype, with false recognition of the category prototype higher in the cross-modality conditions. The results are discussed in terms of current theories in modality transfer including the difficulties inherent when calculation of similarity must be considered in a cross-modal situation.

Category inference as a function of correlational structure, category discriminability, and number of available cues.

Two experiments investigated category inference when categories were composed of correlated or uncorrelated dimensions and the categories overlapped minimally or moderately. When the categories minimally overlapped, the dimensions were strongly correlated with the category label. Following a classification learning phase, subsequent transfer required the selection of either a category label or a feature when one, two, or three features were missing. Experiments 1 and 2 differed primarily in the number of learning blocks prior to transfer. In each experiment, the inference of the category label or category feature was influenced by both dimensional and category correlations, as well as their interaction. The number of cues available at test impacted performance more when the dimensional correlations were zero and category overlap was high. However, a minimal number of cues were sufficient to produce high levels of inference when the dimensions were highly correlated; additional cues had a positive but reduced impact, even when overlap was high. Subjects were generally more accurate in inferring the category label than a category feature regardless of dimensional correlation, category overlap, or number of cues available at test. Whether the category label functioned as a special feature or not was critically dependent upon these embedded correlations, with feature inference driven more strongly by dimensional correlations.

Exploring the psychological structure of transformational knowledge in visual and haptic intramodal conditions using multidimensional scaling.

The intramodal relation between perceptual similarity and categorization performance in a psychological space, as indicated by multidimensional scaling (MDS) analysis of similarity judgments, was explored. Participants learned to classify transformed object shapes into three categories either visually or haptically via different training procedures (either random or systematic), followed by a transfer test. Learning modulated the psychological spaces, but this effect was more prevalent with haptic than with visual tasks. A prototype model for similarity ratings was illustrated in MDS space. The prototypes were multidimensionally scaled at the center of a category, rather than mirroring the bidirectional paths of their origins. Although they converged at the apex of two transformational trajectories, the category prototypes anchored at the centroid of their respective categories and became more structured as a function of learning. The reduced tendency to make errors (i.e., higher accuracy) in recognizing and classifying the category prototypes suggested that prototypical representation of a category abstracted from exemplar averaging functioned more as novel, rather than familiar, information. Findings were discussed in terms of transformational knowledge, categorical representation in three-dimensional (3D) space, and intramodal visual and haptic similarity.

The enduring effect of availability.

The present study addressed whether experienced ease of retrieval of autobiographical events affects trait judgments made immediately and after a delay (Experiment 1) and whether this influence is modulated by either of two discounting instructions (Experiment 2). In Experiment 1, participants first attempted to retrieve 6 or 12 personal memories for different traits and then made ease of retrieval and self-trait judgments immediately or 1 week later. In Experiment 2, participants made immediate ratings and then delayed ratings under 3 instructional manipulations: a generic instruction to repeat these judgments, to make these ratings after they were reminded of their original recall total, or to make these ratings after they had their retrieval ease discounted or augmented. In each experiment, an enduring effect of availability was obtained in that the relationship between ease of retrieval and self-trait rating was only slightly affected by the delay. Being reminded of the original number of recalled memories nullified the relationship between ease of retrieval and trait rating. However, discounting or augmenting ease of retrieval, which altered ease of retrieval ratings, did not. Potential explanations for an enduring effect of availability are discussed.

Attentional costs and benefits in memory search.

The role of attentional mechanisms in memory search was investigated in 2 experiments. In Experiment 1, an abrupt-onset cue was a valid or invalid predictor of a spatially displaced memory probe in a memory search paradigm. The 2 conditions differed only in terms of the duration of the memory probe: either 200 ms or an unlimited duration until the subject's response. We found that memory probe duration had little impact on memory search, as revealed by the slope across memory set size, although an invalid prior cue slowed responding by increasing the intercept by about 70 ms. In Experiment 2, costs and benefits of valid and invalid cues were assessed by inclusion of a neutral condition. Both costs and benefits were found, with effects again localized in the intercept of the memory search functions. A simple model was proposed that estimated 2 attentional transit times, 1 to the abrupt-onset cue and 1 activated after disengagement from an invalid location. We address whether the rapid examination of the contents of working memory should be considered an encapsulated process, unperturbed by abrupt-onset events that delay but do not otherwise disturb the resulting search.

Delayed memory for visual-haptic exploration of familiar objects.

Long-term memory of haptic, visual, and cross-modality information was investigated. In Experiment 1, subjects briefly explored 40 commonplace objects visually or haptically and then received a recognition test with categorically similar foils in the same or the alternative modality both immediately and after 1 week. Recognition was best for visual input and test, with haptic memory still apparent after a week's delay. Recognition was poorest in the cross-modality conditions, with performance on the haptic-visual and visual-haptic cross-modal conditions being nearly identical. Visual and haptic information decayed at similar rates across a week delay. In Experiment 2, subjects simultaneously viewed and handled the same objects, and transfer was tested in a successive cue-modality paradigm. Performance with the visual modality again exceeded that with the haptic modality. Furthermore, initial errors on the haptic test were often corrected when followed by the visual presentation, both immediately and after 1 week. However, visual test errors were corrected by haptic cuing on the immediate test only. These results are discussed in terms of shared information between the haptic and visual modalities, and the ease of transfer between these modalities immediately and after a substantial delay.

Memory search following valid and invalid abrupt-onset cues.

The impact of abrupt-onset cues on memory search was investigated, where the abrupt-onset cue was a valid (Experiment 1), random (Experiment 2), or irrelevant (Experiment 3) predictor of the location containing the test probe. In Experiment 4, the abrupt-onset cue either preceded or followed the test probe. Sternberg-like functions were obtained in Experiments 1 and 2, with the effects of the abrupt-onset cue localized primarily in the intercept rather than the slope. Experiment 3 demonstrated that a spatially separated and irrelevant abrupt-onset cue increased latency even when all memory probes occurred at the fixation point. In Experiment 4, the robust impact of an abrupt-onset cue vanished, regardless of stimulus onset asynchrony, when it followed the target. We concluded that abrupt-onset cues captured attention regardless of their predictability, manifested as a delaying of search. However, once attention was captured by the target, a subsequent abrupt-onset stimulus had no effect. These results were discussed in terms of diffuse attention and contingent capture models of attention.

Memory for temporally dynamic scenes.

Recognition memory was investigated for individual frames extracted from temporally continuous, visually rich film segments of 5-15 min. Participants viewed a short clip from a film in either a coherent or a jumbled order, followed by a recognition test of studied frames. Foils came either from an earlier or a later part of the film (Experiment 1) or from deleted segments selected from random cuts of varying duration (0.5 to 30 s) within the film itself (Experiment 2). When the foils came from an earlier or later part of the film (Experiment 1), recognition was excellent, with the hit rate far exceeding the false-alarm rate (.78 vs. 18). In Experiment 2, recognition was far worse, with the hit rate (.76) exceeding the false-alarm rate only for foils drawn from the longest cuts (15 and 30 s) and matching the false-alarm rate for the 5 s segments. When the foils were drawn from the briefest cuts (0.5 and 1.0 s), the false-alarm rate exceeded the hit rate. Unexpectedly, jumbling had no effect on recognition in either experiment. These results are consistent with the view that memory for complex visually temporal events is excellent, with the integrity unperturbed by disruption of the global structure of the visual stream. Disruption of memory was observed only when foils were drawn from embedded segments of duration less than 5 s, an outcome consistent with the view that memory at these shortest durations are consolidated with expectations drawn from the previous stream.

The indirect modification of categorical knowledge.

The present study investigated whether the later learning of a category could affect the representation of other categories learned previously. Participants initially learned two or three categories, where each stimulus was composed of features that were distinctive to a category, shared with one or both of the other categories, or were idiosyncratic. When two categories were initially learned, a subsequent learning phase involved the learning of a third category that either shared distinctive features with categories learned previously, thereby discounting those features as diagnostic or was composed of features unrelated to the original categories. A common transfer test contained old, new, and prototype stimuli for classification, as well as critical items that revealed whether discounting of previously diagnostic features had occurred. The results revealed that stimuli assigned to a particular category in the two-category condition were assigned to the third category learned subsequently when the later learning discounted previously diagnostic features. These results suggest that later learning of a category can indirectly modify the representation of categories learned previously.

Array training in a categorization task.

Two components of categorization, within-category commonalities and between-category distinctiveness, were investigated in a categorization task. Subjects learned three prototype categories composed of moderately high distortions, by observing arrays containing patterns that belonged either to a common prototype category or to three different categories; a third group learned patterns presented one at a time, mirroring the standard paradigm. Following 6 learning blocks, subjects transferred to old patterns and new patterns at low-, medium-, and high-level distortions of the category prototype. The results showed that array training facilitated learning, especially when patterns in the array belonged to the same category. Transfer results showed a strong gradient effect across pattern distortion level for all conditions, with the highest performance obtained following array training on different category patterns and worst in the control condition. Interestingly, the old training patterns were classified worse than new low and no better than medium distortions. Neither this ordering nor the steepness of the gradient across prototype similarity for each condition could be predicted by the generalized context model. A prototype model better captured the steep gradient and ordinal pattern of results, although the overall fits were only slightly better than the exemplar model. The crucial role played by category commonalities and distinctiveness on categorical representations is addressed.

Bogus concerns about the false prototype enhancement effect.

Two experiments addressed the mechanism responsible for the false prototype effect, the phenomenon in which a prototype gradient can be obtained in the absence of learning. Previous demonstrations of this effect have occurred solely in a single-category paradigm in which transfer patterns are assigned or not to the learning category. We tested the hypothesis that any extraneous variable potentially responsible for this effect, such as compactness varying with pattern distortion (Zaki & Nosofsky, 2004), may be functional in the single-category paradigm but not when multiple categories are available at the time of transfer. In the present study, subjects received a bogus or a real category learning phase, followed by a transfer test that required assignment into 1 or 3 prototype categories. The results showed that a minimal prototype gradient was obtained in the bogus conditions, with performance approaching chance levels when classification into 3 categories was required. In contrast, a substantial prototype gradient effect was found following learning. We conclude that the prototype gradient typically obtained following multiple-category learning is primarily driven by real learning and that the false prototype effect is itself an artifact of the single-category paradigm.

Haptic concepts in the blind.

We investigated and compared the acquisition of haptic concepts by the blind with the acquisition of haptic concepts by sighted controls. Each subject--blind, sighted but blindfolded, sighted and touching, and sighted only--initially classified eight objects into two categories using a study/test format, followed by a recognition/classification test involving old, new, and prototype forms. Each object varied along the dimensions of shape, size, and texture, with each dimension having five values. The categories were linearly separable in three dimensions, but no single dimension permitted 100% accurate classification. The results revealed that blind subjects learned the categories quickly and comparably with sighted controls. On the classification test, all groups performed equivalently, with the category prototype classified more accurately than the old or new stimuli. The blind subjects differed from the other subjects on the recognition test in two ways: They were least likely to false alarm to novel patterns that belonged to the category but most likely to false alarm to the category prototype, which they falsely called "old" 100% of the time. We discuss these results in terms of current views of categorization.

The modulating influence of category size on the classification of exception patterns.

Generalization gradients to exception patterns and the category prototype were investigated in two experiments. In Experiment 1, participants first learned categories of large size that contained a single exception pattern, followed by a transfer test containing new instances that had a manipulated similarity relationship to the exception or a nonexception training pattern as well as distortions of the prototype. The results demonstrated transfer gradients tracked the prototype category rather than the feedback category of the exception category. In Experiment 2, transfer performance was investigated for categories varying in size (5, 10, 20), partially crossed with the number of exception patterns (1, 2, 4). Here, the generalization gradients tracked the feedback category of the training instance when category size was small but tracked the prototype category when category size was large. The benefits of increased category size still emerged, even with proportionality of exception patterns held constant. These, and other outcomes, were consistent with a mixed model of classification, in which exemplar influences were dominant with small-sized categories and/or high error rates, and prototype influences were dominant with larger sized categories.