Individual Submission Summary
Share...

Direct link:

Poster #7 - The differential influence of comparison on children's word learning with an imposed delay

Sat, March 23, 9:45 to 11:00am, Baltimore Convention Center, Floor: Level 1, Exhibit Hall B

Integrative Statement

Multiple studies demonstrate benefits of comparison and contrast for learning a variety of higher level relations. We examined the effects of comparison and contrast with learning non-relational perceptual information, specifically on 3-year-old’s learning of labels for novel shape categories. In Study 1 children were tested immediately and in Study 2 children were tested after a delay. In both studies novel words were presented in three between-subject conditions: comparison, contrast, and one-example. Figure 1 shows the 3 learning conditions and test objects.
Learning Phase:
Comparison: all objects were the same shape, but different colors and textures. Each object was labeled with a novel word (e.g., “This is a wug”). Contrast: one object was the target shape, and two objects differed from the target on shape, texture, and color. The target shape was labeled “This is a wug”, and the non-targets were labeled “This is not a wug”. One-example: one object of the target shape category labeled “This is a wug” three times.
Test Phase:
There were four objects, a shape match, a novel object with a color and texture seen in the learning phase, and two distracter objects. Children were asked to extend the label to a new example of the category. As Figure 2 shows, in Study 1 learning from one example outperforms learning from multiple examples via comparison or contrast, counter-intuitively suggesting a detrimental role of comparison and contrast for non-relational categories. The results of Study 2 did not follow the same pattern suggesting that one example is enough for immediate recognition, but not when children need to consolidate and recall the category after a delay. This is consistent with previous research that finds comparison with spacing between instances of learning is better than simultaneous comparison when there is a delay between learning and test.

Authors