Can toddlers learn causal action sequences?

• In Event: Poster Session 10
  In Poster Session: PS 10 Topic - Cognitive Processes

Abstract

The ability to identify causal relationships is a fundamental cognitive skill. Like older children and adults, toddlers can learn simple cause-effect relationships between a single action and its outcome—for example, that pressing a button on a puzzle-box causes it to dispense a sticker. However, successfully navigating the real world requires understanding of more complex causal relationships.

One useful way to figure out how artefacts work is to watch other people interacting with them. In previous work (Anonymized), children watched an intentional adult demonstrator perform a 2-action sequence (AB) on a puzzle-box (e.g., turning a knob then pressing a button) that then dispensed a sticker (effect E). Although only the second action (B) was causally necessary for bringing about the effect, it was ambiguous whether the first action (A) was also required. When given the opportunity to interact with the box, preschoolers tended to reproduce the 2-action sequence they had seen demonstrated (AB), whereas toddlers typically only performed the second action from the sequence (B). Subsequent work showed that toddlers do not simply act on the basis of a recency effect in this paradigm—when they saw an A-E-B demonstration, so that the first action A preceded the sticker being dispensed and the second action B followed it, they correctly acted on A (Anonymized). These findings raise the question of whether toddlers have difficulty learning causal action sequences—e.g., that you need to first turn a knob and then press a button to get a sticker.

To explore the development of children’s ability to understand causal action sequences, Experiment 1 of the present study asked whether children can learn that a sequence of two actions is necessary to produce an effect. In Experiment 1a, 12- to 35-month-olds (N=88) saw evidence that a 2-action sequence (AB) led to an effect, whereas B alone did not (Figure 1A). When given the opportunity to activate the puzzle-box up to five times, 48.9% of toddlers performed the ineffective action B first on their first attempt. Younger toddlers in particular found it challenging to produce the sequence required to get a sticker—1-year-olds only activated the puzzle-box an average of 3.07±0.32 times out of five, compared with 4.23±0.23 times for 2-year-olds. Experiment 1b is investigating how preschoolers (36- to 71-month-olds, N=58, ongoing) perform in the same task. Preliminary results of Experiment 1 suggest the ability to infer that a 2-action sequence is causally necessary increases across early childhood (rs(144)=0.61, p<0.001, Figure 2). Experiment 2 is investigating what causal inferences toddlers (N=37, ongoing) and preschoolers (ongoing) make when they are shown that both AB and B alone produce the same effect (Figure 1B). Of interest is whether they grasp that a sequence is not necessary based on this evidence.

Our preliminary findings suggest that the ability to infer that a sequence of two actions is causally necessary may develop over early childhood, thus adding to our understanding of children’s ability to infer causal structure from the actions they see others perform.

Authors

• Emma C Tecwyn, Birmingham City University
  Non-Presenting Author

• Nishat Tasnim Kazi, University of Toronto
  Presenting Author

• Nafisa Mahbub, University of Toronto
  Non-Presenting Author

• Daphna Buchsbaum, Brown University
  Non-Presenting Author