The prediction of emotional decision making from working memory and inhibitory control in preschool children: using decision tree model

Abstract

ABSTRACT While there is a theoretical distinction between cool (cognitive) and hot (emotional) executive functions, potential relationships can be identified between tasks associated with these two aspects. A decision tree serves as a model for training and analyzing data, predicting the target variable based on independent variables in a hierarchical fashion. In contrast to other predictive methods, this model doesn’t necessitate statistical expertise and makes decisions akin to human decision-making through hierarchical “if” and “then” rules, providing an easily interpretable framework. To date, no studies have explored the relationship between cool and hot executive functions using decision tree models. In this study, preschool children specifically Persian-speaking Iranian children aged 4 to 5 years (N = 71, M age = 59.07; SD = 6.03), participated in cool (Forward Digit Span, Backward Digit Span, Day-Night inhibitory control, and Happy-Sad Inhibitory control) and hot (Children Gambling Task) executive function tasks. Analyses were performed using MATLAB programming software. The C4.5 version of the decision tree was employed to predict the final CGT blocks’ scores using scores from cool executive function tasks as inputs. By employing this method, a minimal prediction error (approaching zero) was achieved, significantly showing the robust predictive capability of cool executive function in anticipating hot executive function. This outcome suggests potential relationships between the cognitive and emotional aspects of executive function.