Intuitive Statistics & Metacognition in Children and Adults
- Madeline Pelz, MIT, Cambridge, Massachusetts, United States
- Laura Schulz, Brain and Cognitive Sciences, MIT, Cambridge, Massachusetts, United States
AbstractAcross four experiments, we look at whether adults and children can represent the amount of information needed to distinguish different populations in the context of an intuitive statistical reasoning task requiring metacognitive monitoring and control. Adults (N=60) modulated their information gathering with respect to the difficulty of the discrimination problem consistent with a ground truth model of information gain. Adults also adjusted their confidence threshold depending on task difficulty, allowing for more uncertain judgments when the discrimination was more difficult or gathering data was more costly (Experiments 1 and 2). In a simplified version of the task, children (N = 42, M = 7.3 years, range: 5.0-9.0) were also able to distinguish easy and difficult discrimination problems and judge that they needed more information to solve harder problems (Experiments 3 and 4).
Return to previous page