Development of Non-Symbolic Probabilistic Judgment in Children: Intuitive Biases and Individual Differences

Research Hypothesis
The study hypothesized that children's performance in non-symbolic probabilistic judgment tasks would improve with age, but intuitive biases would persist, especially in incongruent tasks where more targets lead to lower probability.
Data and Methodology
Participants
The study involved 415 elementary school students from grades 2 to 5 in Wuhan, China, with ages ranging from 7.37 to 10.34 years.
Probability Judgment Task
Participants completed a paper-and-pencil task with 24 trials (12 congruent and 12 incongruent) involving choosing the box with the highest probability of drawing a target ball. The order of trials was pseudo-randomized.
Key Findings
Accuracy and Intuitive Biases
Children performed better on congruent trials than incongruent trials across all grades.
Intuitive biases did not decrease with age; in fact, fifth graders showed the largest bias.
A significant interaction between grade level and task type was observed, but overall accuracy did not improve with age.
Latent Class Analysis (LCA)
LCA identified four subgroups:
High Probability Judgment Ability : High accuracy, low biases.
Moderate Ability, Easily Influenced by Intuition: Moderate accuracy with significant intuitive biases.
Low Ability, High Reversal Bias: Low accuracy with pronounced reversal biases.
Low Ability, Easily Influenced by Intuition: High accuracy on congruent trials, but significant intuitive biases on incongruent trials.
Interpretation
The findings challenge the assumption that probabilistic reasoning improves automatically with age. Instead, intuitive biases persist and may increase, highlighting the need for targeted educational interventions to address these biases.
Practical Implications
The study suggests that educators should design curricula that emphasize proportional reasoning and strategies to overcome heuristic thinking. Future research should explore longitudinal designs and cross-cultural comparisons to validate these findings and uncover underlying cognitive mechanisms.