The effect of training methodology on knowledge representation in categorization

Abstract

Category representations can be broadly classified as containing within-category information or between-category information. Although such representational differences can have a profound impact on decision-making, relatively little is known about the factors contributing to the development and generalizability of different types of category representations. These issues are addressed by investigating the impact of training methodology and category structures using a traditional empirical approach as well as the novel adaptation of computational modeling techniques from the machine learning literature. Experiment 1 focused on rule-based (RB) category structures thought to promote between-category representations. Participants learned two sets of two categories during training and were subsequently tested on a novel categorization problem using the training categories. Classification training resulted in a bias toward between-category representations whereas concept training resulted in a bias toward within-category representations. Experiment 2 focused on information-integration (II) category structures thought to promote within-category representations. With II structures, there was a bias toward within-category representations regardless of training methodology. Furthermore, in both experiments, computational modeling suggests that only within-category representations could support generalization during the test phase. These data suggest that within-category representations may be dominant and more robust for supporting the reconfiguration of current knowledge to support generalization.

Fig 1. Stimuli used in Experiment 1.

The x-axis corresponds to the width of the bars and the y-axis corresponds to the rotation angle of the bars. Symbols denote different categories. The mean stimulus of each category is shown as an example.

Fig 2. Experimental procedures used in both experiments.

The top line shows an example A/B training trial while the bottom line shows an example Yes/No training trial. Test trials were identical except that feedback was omitted.

Fig 3. Mean accuracy per block in Experiment 1.

Blocks 1-5 are the training phase and Block 6 is the test phase. Error bars are between–subject standard error of the mean.

Fig 4. Mean accuracy per block in Experiment 1 separated by best–fitting model.

(a) Density–based model. (b) Boundary–based model. Error bars are between–subject standard error of the mean.

Fig 5. Stimuli used in Experiment 2.

The x-axis corresponds to the width of the bars and the y-axis corresponds to the rotation angle of the bars. Symbols denote different categories. The mean stimulus of each category is shown as an example.

Fig 6. Mean accuracy per block in Experiment 2.

Blocks 1–5 are the training phase and Block 6 is the test phase. Error bars are between–subject standard error of the mean.

Fig 7. Mean accuracy per block in Experiment 2 separated by best–fitting model.

(a) Density–based models. (b) Boundary–based models. Error bars are between–subject standard error of the mean.