Are children mindful decision makers? Investigating underlying mechanisms of Turkish-learning children's recency bias

This study examined (a) whether Turkish-reared preschoolers (n = 59, Mage = 53.37 months, SD = 10.11) showed recency bias in a flexible word order language, and (b) how their recency bias was related to individual differences in working memory and verbal skills. Children were presented with a decision-making task in which they responded to questions about an imaginary character's decisions. They listened to questions that had both choice options at the beginning, at the end or first option being at the beginning while the second at the end of the sentence. Then, children were given tests to assess their working memory (WM) and verbal skills. We found that children exhibited recency bias, and it decreased as their age increased. Children's WM skills, not language skills, were a strong predictor of recency bias, even after controlling for age. Whether the options were familiar or located at different places within the question did not have an effect alone. However, slope analyses showed that their interactions with each other and children's WM skills may have yielded a significant decrease in recency bias. That is, as children aged, their recency bias decreased when responding to unfamiliar options if both were encountered at the beginning of the question. Recency bias decreased as their WM skills improved when both options were encountered at the beginning of the questions. This study investigated recency bias systematically in a free word order language with respect to individual differences, suggesting that children are not always mindful decision-makers. These findings have implications for unveiling children's response biases and ways to elicit effective communication with them. Keywords: recency bias, working memory, response bias, decision making, Turkish word order