The effect of dual-task cost on strategy execution in arithmetic with event-related potentials.

The present study adopted a dual-task paradigm manipulating operand presentation sequence (continuous vs. discontinuous) and a no-choice strategy adoption approach to investigate how stimulus presentation and response-related processes modulate strategy execution in dual-task scenarios. Participants completed a series of two-digit addition computational estimation tasks, in which they were required to use either the rounding-down (RD) strategy (e.g., calculating 50 + 80=130 for the problem 54 + 89) or the rounding-up (RU) strategy (e.g., calculating 60 + 90=150 for 54 + 89). These tasks were administered across both single- and dual-task conditions, with operands presented in either a continuous or discontinuous sequence. A critical dissociation emerged between neural and behavioral outcomes: significant interaction effects among presentation sequence, task condition, and strategy type were detected at the neural level, whereas no such interactions were observed in behavioral metrics. Behaviorally, relative to the single-task condition, RD strategy, and continuous presentation sequence (MMC first), the dual-task condition, RU strategy, and discontinuous presentation sequence (CE first) were associated with significantly lower accuracy (ACC), longer reaction times (RTs), and higher Inverse Efficiency Scores (IES). At the neural level, discontinuous operand presentation in the dual-task condition elicited significantly larger N1 and P2 amplitudes over the parietal lobe, reflecting heightened attentional engagement compared to continuous presentation. A striking cross-condition reversal was observed for N2 and P3 amplitudes: in the single-task condition, N2 and P3 amplitudes under the RD strategy were significantly smaller (more negative) than those under the RU strategy; conversely, in the dual-task condition, RD strategy-related N2 and P3 amplitudes were significantly larger (more positive) than those linked to the RU strategy. These findings align closely with the Time-Based Resource-Sharing (TBRS) model and dynamic processing frameworks. Collectively, the results indicate that sequence preparation, dual-task coordination, and strategy execution depend on attention-dependent temporal allocation of limited cognitive resources at central processing bottlenecks, which enables flexible, optimal, and dynamic resource management to support both online task processing and cognitive maintenance. Future investigations into the neural mechanisms of dual-task processing should employ larger sample sizes and multimodal research approaches to further elucidate these resource allocation dynamics.