Young Children are Sensitive to their Learning Curve
Persistence is critical to academic achievement. Yet, there is not much known about how young children make decisions about when and how to persist. Therefore, we don't know how to help children persist through challenges when it matters most. Adult models of optimal learning indicate that adults are capable at monitoring their past performance over time and use this information to determine where to put their effort. However, it is unknown if children are capable of this too. Here, we explored whether 4-6-year-olds track their past performance over time and use this information to determine when to stick with a challenge. Across four experiments (N=360), we found that children are sensitive to the trajectory of their past performance over time and use this information to determine if they should stick with a challenge. Children are more likely to continue with a challenge when their performance increases over time rather than stays the same. Additionally, reward contingencies shift preferences. Children integrate their chances of getting a reward with the magnitude of that reward to calibrate their effort. With in-person testing, older children are more likely to stick with a challenge and are less optimistic about their performance than younger children regardless of their performance. However, these findings are not found in online testing. Furthermore, older children are more accurate at updating performance predictions than younger children across in-person and online contexts.
Comments
Really nice study and super…
Really nice study and super creative adaptation of paradigm to online task. Do you have any idea why the age effects disappeared in the online condition? Do all children think they're equally adept? I still worry just a little about the perceived control that kids had. Your results certainly indicate that the kids believed they were getting better outcomes, (and were therefore more willing to persist) but did they ACTUALLY believe they were "learning"? Learning means they were acquiring a skill - that they were themselves getting better at the task, and that wasn't actually the case here. Nothing they did made any difference to the actual outcome, so this wasn't really a learning curve - it was really a luck curve. Can you think of a way to design a study that would reflect the actual acquisition of skill? It's tricky to control things in a study like that, but that would be a neat next step in this line of research!
Impressive line of research. I hope you keep it up!
Cheers,
Dr. Melissa Hunt
Hi Skyler, Very clear and…
Hi Skyler,
Very clear and professional presentation. You should be proud of the work you did. And it will lay the foundation for potentially more fundamental work. As you have undoubtedly discussed with Allyson, it is sometimes difficult to know whether these laboratory tasks teach us about how things work in the real world. I wonder whether you think it would be possible to develop an experiment designed around a task with more real-world relevance, where the structure of the task is similar but in which the rewards may have more visceral value to the child (praise from a parent or an authority figure, cookies, screen-time, etc.). I have a toddler and I have been struck by how well he tracks what questions, tone of voice, and general strategies writ large get his parents to do what he wants (e.g. go to the park or the basketball court). Developing tasks more tightly connected to the real world is difficult and not always possible, but is certainly worth considering.
Good job!
Best,
Johannes