Infants rationally decide when and how to deploy effort

• In Event: Understanding the origins, determinants, and consequences of children’s mastery motivation: A developmental approach

Abstract

Persistence in the face of obstacles is a double-edged sword. On one hand, it is central to heightened academic achievement and emotional well-being (1-2); on the other hand, it is metabolically expensive and rife with opportunity costs. What truly spurs learning is adaptive persistence: deciding whether, when and how to try based on the perceived likelihood of one’s success at the task. We investigated whether infants adaptively vary the nature of their persistence by using both the social input they receive (that a problem is easy, difficult, or impossible to solve), and ongoing first-person experience with the task (3).
In a between-subjects experiment, we gave 18-month-olds (N = 96) a task in which a rope attached to an out-of-reach box could be pulled to get a toy in the box (Fig 1). Prior to the task, infants learned that an adult experimenter could either easily solve the problem (easy condition, n = 32), found the problem challenging but solvable (hard condition, n = 32), or was not able to solve the problem (impossible condition, n = 32). We then examined infants’ persistence on the same task, which unbeknownst to them was actually impossible to solve (as the container was glued to the table) on three test trials.
Persistence consists of both a temporal component (deciding how long to try), and an intensity component (deciding how hard to try); thus we measured both infants’ time trying and their pulling force (via a hidden force gauge). This combination of measures provides insight into the expectations infants hold about the solvability and difficulty of the task. Infants’ time spent trying across trials varied as a function of condition (Fig 2). In the impossible and easy conditions, trying time decreased dramatically across trials: t(63) = -4.56, p < .001 and t(63) = -5.94, p < .001, respectively. In contrast, in the hard condition, infants’ trying time remained relatively constant across time: t(63) = -1.96, p = .05. Infants’ maximum pulling force across trials also interacted with condition: Infants exerted consistently low levels of force in the impossible condition: t(46.9) = .11, p = .91, consistently high levels of force in the easy condition: t(46.35) = -1.57, p = .12, and increasingly high levels of force in the hard condition: t(50.67) = 2.99, p = .004. Thus, infants use social input to form expectations of task difficulty, which are then revised in light of their own experience, and subsequently guide when and how infants try.
We also assessed infants’ (1) affect to probe expectations about how hard/easy the task should be, (2) help-seeking behaviors to gauge their inferences about the experimenter’s ability to solve the task; and (3) performance during a “recovery trial”, when the task changed and possible to solve, to assess downstream motivational consequences.
Overall, our results demonstrate that infants are rational decision makers: they dynamically integrate social evidence and their own experience to decide when, whether and how to deploy effort. This ability likely underlies infants’ capacity for rapid and extraordinary learning.

Authors

• Kelsey Lucca, Arizona State University
  Presenting Author

• Rachel Horton, University Of Washington
  Non-Presenting Author

• Jessica A Sommerville, University of Toronto
  Non-Presenting Author