Caregivers Facilitate Category Learning in Infants through Labeling

• In Event: Poster Session 03
  In Poster Session: PS 03 Topic - Cognitive Processes

Abstract

Categories provide a framework as infants begin to learn about their environment by affording inferences about novel exemplars and supporting memory and retrieval of information (Bornstein & Arterberry, 2010). The processes underlying category formation are still being understood. Research on category formation in infancy has mostly employed highly controlled visual fixation paradigms in which instances of novel categories are systematically presented (e.g., Waxman & Markow, 1995). But how are categories formed in the real world, during interactions between infants and parents? From the literature, we hypothesized that infants would learn categories better in a dyadic than in a solo condition given that the dyadic condition allows for behaviors such as labeling to occur and support category formation.
In a within-subjects design, infants (N = 59, Mage = 12.11 months, range = 11.00-14.13 months) played with four different sets of toys; two real category sets (biological: fruits, animals; artifactual: tools, vehicles) and two invented category sets (‘sprocks’ distinguished by paws and feet or ‘dills’ distinguished by faces) either in a dyadic play session with their caregiver or a solo play session. In dyadic and in solo play, infants were given one real set and one invented set. Invented categories were included to reveal the importance of category labels as infants had no prior experience with these categories. We predicted that dyadic play would lead to better category formation than solo play, as measured by a 10-second visual fixation period examining novelty preference after each play session and that mothers’ production of labels would assist category formation.
No significant main effects of dyadic versus solo play emerged; however, differences in caregivers’ use of labels related to category formation. Probing caregiver labeling revealed that for artifactual (but not biological) categories there was a correlation between label use and novelty preference (rs = .52, p = .003). Similarly, a relation between label use and novelty preference was found for dills (rs = .49, p = .02) but not for sprocks (rs = -.10, p = .66). Though there were no differences in caregivers’ use of category labels (U = 398.50, z = -.27, p = .78), caregivers used individual exemplar labels (e.g., dog) more often with biological categories than artifactual categories (U = 163.50, z = -4.20, p = .000). Individual exemplar labels may have interfered with infants’ learning of category labels for biological toys.
This is the first study to probe how infants form categories in naturalistic interactions with parents. Findings have implications for evaluating the importance of talking with children in ways that promote category formation. Future research should continue to explore the role of labels and additional behaviors, such as object motion, in category formation during parent-child play sessions. Thus, it is not dyadic play per se that promotes category formation but whether parents engage in category labeling.

Authors

• Hannah Puttre, Boston University
  Presenting Author

• David Neal, University of Cambridge
  Non-Presenting Author

• Alexa Bruette, University of Delaware
  Non-Presenting Author

• Kathleen H. Corriveau, Boston University
  Non-Presenting Author

• Roberta M Golinkoff, University of Delaware
  Non-Presenting Author

• Kathy A Hirsh-Pasek, Temple University
  Non-Presenting Author