New perspectives to understand individual differences: application of the drift diffusion model to a numerical stroop task in children

The Drift Diffusion Model (DDM) is a computational framework used in cognitive psychology research to analyze the processes underlying two-option decisions. According to the DDM assumption, individuals accumulate evidence until reaching a threshold that favors one possible response over another. This study applied the DDM to investigate how children at different developmental stages (elementary vs. middle school) respond to a Numerical Stroop Task. The DDM parameters estimation was performed using a Bayesian Hierarchical approach. The estimated parameters included the drift rate (v), which represents the information accumulation rate, the boundary threshold (a), which indicates the amount of information required to select a response, and the non-decision time (t0), which encompasses processes unrelated to the decision-making itself, such as sensory encoding and response execution. Results showed that both younger and older children performed better on congruent and worse on incongruent trials, indicated by higher and lower drift rate, respectively, as compared to neutral condition. This suggests that congruent trials enhance performance with extra information (i.e., number size), while incongruent trials produced an interference effect. Furthermore, the significantly lower drift rate observed in the younger group reflects an overall worse performance, potentially due to the extraction of lower-quality data by younger than older children. Moreover, the difference between groups in t0 suggests that other latent task-related processes, such as motor skills, could affect performance. These findings highlight how the application of DDM - integrating both accuracy and RTs- might be a complementary tool in analyzing the process underlying developmental differences.