Smart or just lucky? Inferring question-asking competence from strategies’ efficiency versus effectiveness.

Abstract

Previous research shows that children evaluate the competence of others based on how effectively someone accomplished a goal, that is, based on the observed outcome of an action (e.g., number of attempts needed). Here, we investigate whether 5-to 10-year-old children and adults infer competence from how efficiently someone solves a task by implementing question-asking strategies of varying expected information gains (EIG). Whereas the efficiency of a strategy defined as EIG is a reliable indicator of competence, the observed effectiveness of actions may depend on unrelated external factors, such as luck. Across two experiments conducted in Germany, we varied how efficiently and how effectively different agents solved a 20-questions game (Experiments 1 and 2) and a maze-exploration game (Experiment 2). In Experiment 1 (N = 121), only adults identified a more efficient agent as more competent, and all participants attributed higher competence to agents needing fewer questions even when they employed the same inefficient strategy. In Experiment 2 (N = 220), adults and children from about 8 years onward successfully identified the agents using the more efficient strategy as more competent. Overall, our results suggest that observed effectiveness is a powerful cue for competence even when such an inference may not be warranted and that the ability to make explicit competence judgments based on the efficiency of a strategy alone emerges around 8 years of age, although, as shown in previous work, a more implicit understanding of competence may already be present during the preschool years. Public Significance StatementThis study investigated the cues young children and adults use to evaluate the competence of someone they observed solving a problem. Children struggle to evaluate competence based on the efficiency of the strategy used until around age 8. Instead, the observed number of steps taken to reach a solution remains a powerful cue for competence throughout childhood and even adulthood. Children's accuracy in making inferences about other people's competence is relevant for their skill growth, socio-cognitive development, and academic achievement. Therefore, exploring and promoting an unbiased understanding of competence, stressing the importance of strategy efficiency, can support children's development in classroom settings and beyond.

Children's information‐search strategies: Operationalizing efficiency and effectiveness

“…Therefore, what we call luck is the variability in effectiveness for a fixed EIG value (see Figure 1b). Because of luck, inferences about an agent's information-search competence are accurate only if they are based on the agent's efficiency rather than on their observed, lucksensitive effectiveness (Török et al, 2023). Nevertheless, we acknowledge that tracking the effectiveness of queries may serve as a useful heuristic to infer question-asking competence.…”

“…This is because the predicted behavioral outcomes based on efficiency and effectiveness often correlate, and observing cues to effectiveness may be less costly computationally and faster than computing and tracking efficiency based on EIG. Indeed, in one study (Török et al, 2023), children younger than 8 years old were more likely to use the difference in the number of questions asked by two agents to evaluate their competences, whereas older children relied more on EIG differences. More work is needed to investigate children's and adults' potential use of such heuristics, as well as their ability to go beyond them.…”

Seeking new information with old questions: Children and adults reuse and recombine concepts from prior questions

“…In contrast, in the previously-uninformative condition, the question was less informative and the confederate made an incorrect guess. We manipulated both informativeness and guess accuracy because prior research suggests that children sometimes attend to success over informativeness (Török et al, 2023).…”