Kayla Good

Kayla Good

Data analyst & behavioral research scientist | PhD in Psychology (Stanford) | she/her 🌈

The developmental origins of reputation management

At some point or another, we’ve probably all been concerned with wanting others to think we are ‘smart’. As adults, we can recognize how having this kind of reputational concern could influence behavior: someone motivated to appear smart might be less likely to seek out help or be honest about failing at something, since these actions could make others think they’re less competent.

In this project, which I conducted in collaboration with Dr. Alex Shaw at the University of Chicago, we explored when children begin to engage in this kind of reasoning. That is, at what age do children begin to recognize how a desire to appear smart might shape how someone acts around other people?

Across five studies, we recruited a total of 576 four- to nine-year-old children to answer this question. We specifically chose this age range because prior research suggests that this is when children themselves begin engaging in some forms of reputation management, shifting their behavior in response to different audiences, etc. However, what we wanted to know was whether children in this age range have an explicit understanding of this process - i.e., do they predict that someone with reputational motives will behave differently than someone without such motives?

We investigated this by telling our child participants a story about two kids: an intrinsically motivated kid who cares about being smart, and a reputationally motivated kid who cares about appearing smart to others. We then asked them to make predictions about which story character they thought would engage in a particular behavior (e.g., lying to a classmate about having done poorly on a test).

We found that, with age, children predicted the reputationally motivated child would be more likely to lie to cover up failure (p < .001) but less likely to downplay success (p = .003; see below).

Predicted probability of choosing the reputationally motivated student as the one who either lied about poor performance (Study 1, in red) or downplayed success (Study 2, in blue) by age (continuous). Points reflect the individual data for each study; dots at y = 1 indicate choosing the student with reputational concerns, and dots at y = 0 indicate choosing the student with intrinsic concerns. Shaded regions indicate 95% confidence intervals.

We also found that, with age, children thought that the reputationally motivated child would be less likely to seek help in public (where classmates could see them) than in private (p < .001).

Predicted probability of choosing the reputationally motivated student as the one who either sought help in private (in red) or public (in blue) by age (continuous). Points reflect the individual data for each condition; dots at y = 1 indicate choosing the student with reputational concerns, and dots at y = 0 indicate choosing the student with intrinsic concerns. Shaded regions indicate 95% confidence intervals.

Together, these findings suggest that, with age, children are better able to recognize how reputational concerns about appearing smart to others might influence someone’s behavior. It was around age 7 in particular that most of these predictions began to emerge, which suggests that, by the first grade, children know what kinds of actions (e.g., seeking help publicly) might make people think they are less smart. We discussed the potential implications of these findings for education in a recent Scientific American piece.

You can also find a full-write up of these results in our Child Development paper.