Abstract:To be efficient, problem solvers need to be adaptive, tailoring their information search to the specific problem at hand. Across two studies, we investigated the emergence and early development of children's ability to adapt their information search to a given goal (Studies 1 and 2) and to the statistical structure of the problem space (Study 2) to maximize effectiveness. In Study 1, 3-6-year-olds (n = 105) decided which of two cues to look up, the arms or the legs of two monsters, to predict the winner of a throwing or jumping challenge, knowing that monsters with long arms were good throwers and those with long legs were good jumpers. Children's ability to adaptively select relevant information and tailor their search to the goal increased with age, surpassing chance level between the ages of 4 and 5. Study 2 (7-14-year-olds and adults, n = 175) demonstrated this competence in a more complex task, additionally investigating whether children tailor their search to the statistical structure, that is, the distribution of cue values, in their search environment (e.g., how common long legs are). The results suggest high reliability in ignoring irrelevant cues (confirming the results from Study 1) and developmental patterns in children's preferential treatment of cues of differing statistical frequency. Together, these studies contribute to the literature on information search adaptiveness by tracing for the first time the emergence and developmental trajectory of children's ability to tailor predecisional search to the changing goals and environmental resources of the problem at hand. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Parsing heuristic and forward search in first-graders' game-play behavior

Learning Search-Space Specific Heuristics Using Neural Networks

Divide-and-Conquer Monte Carlo Tree Search For Goal-Directed Planning

Case-Based Subgoaling in Real-Time Heuristic Search for Video Game Pathfinding

Dyna-H: a heuristic planning reinforcement learning algorithm applied to role-playing-game strategy decision systems

Metacognitive control during problem solving at early ages in programming tasks using a floor robot

Preschoolers' information search strategies: Inefficient but adaptive

Children adapt their information search to goal and statistical structure of a problem

Life, uh, Finds a Way: Systematic Neural Search

Fast and Precise: Adjusting Planning Horizon with Adaptive Subgoal Search

Rearrangement Planning via Heuristic Search

Scalable Online Planning via Reinforcement Learning Fine-Tuning

Improving Human Decision-Making by Discovering Efficient Strategies for Hierarchical Planning

Adaptive planning depth in human problem solving

Goal-Space Planning with Subgoal Models

State Machine of Thoughts: Leveraging Past Reasoning Trajectories for Enhancing Problem Solving

Monte Carlo Tree Search as a Tool for Self-Learning and Teaching People to Play Complete Information Board Games

Solving Sequential Manipulation Puzzles by Finding Easier Subproblems

A Computational Cognitive Model for Simulating Heuristics Retrieval in Human Problem Solving

People use fast, goal-directed simulation to reason about novel games

Plan of Thoughts: Heuristic-Guided Problem Solving with Large Language Models