The effects of observational comparison and active motor training on infants’ action understanding

• In Event: Object-based understanding and learning: Roles of infant active experience and skill

Abstract

Within the first year of life, infants develop an understanding of other people’s intentions and are constantly anticipating the actions of those around them. At a young age, active experience facilitates action understanding over and above matched observational experience (e.g. Gerson & Woodward, 2015). However, when necessary cues for comparison are present, infants can also learn about new actions through observation (e.g. Gerson & Woodward, 2014). By comparing actions that they cannot perform with actions they can self-produce, infants could develop an intentional understanding of completely novel actions (Gerson & Woodward, 2012). For example, an infant could draw upon their experience with a familiar grasping action (e.g. grasp a toy car by hand) to understand a tool-reach action (e.g. grasp a toy car using a magnet) when presented with a context that encourages comparison of these actions. Experience from comparison could enable infants to generalize relational knowledge to understand different means to achieve the same goal, or novel actions involving a different goal (e.g. grasp a teddy using a claw).
This research examined whether observational experience between hand- and tool-reach comparisons supports 9- to 11-month old’s (Mage=10.07, SD=0.69) predictions of a novel action that they have never produced. In our comparison intervention (N=22), infants observed a hand-reach followed by tool-reach comparisons (Figure 1a), whereas infants in a control condition (N=18) viewed the tool-reaches before the hand-reach. By seeing the hand-reach last, they did not have the hand as a basis for comparison with the tool-reach examples (Figure 1b). In test trials, all infants were required to generalize knowledge from training to a novel context. Infants were familiarized with an agent using a claw to reach for one of two toys (Figure 1c). The toys then swapped positions and visual predictions to the familiarized object were measured using eye-tracking (Figure 1d). After completing the eye-tracking task, we measured infants’ existing ability to actively perform the cloth-reach action viewed during observational training as well as improvement in planful actions after a short active training session.
GLZMs with proportion of accurate visual predictions as the outcome measure tested for the roles of condition, motor ability (existing ability or improvement), and age as covariate. There were no significant main effects of, or interactions between, condition and existing motor ability. However, a significant main effect of improvement in planful actions from pre- to post-training emerged. Follow-up analyses revealed that infants who showed more improvement in planful actions made a higher proportion of accurate predictions. We found no direct support for a unique benefit of comparison on action understanding. It may be that additional cues to compare (e.g. goal labeling) would better facilitate 9- to 11-month-olds’ learning from comparison. Nevertheless, we can conclude that individual differences in infants’ propensity to learn actions facilitates their ability to generalize this knowledge when making sense of novel observed actions. These findings suggest that infants’ motor ability serves as a basis for observational learning, highlighting an important link between active and observational learning in the development of action understanding.

Authors

• Charlotte De Moor, Cardiff University
  Presenting Author

• Sarah A Gerson, Cardiff University
  Non-Presenting Author

• Ross E Vanderwert, Cardiff University
  Non-Presenting Author