Résumé
We investigated the effects of negative emotions on arithmetic performance using ERPs. Participants were asked to verify complex multiplication problems that were either true (e.g., 3 × 23 = 69) or false (e.g., 5 × 98 = 485). Half the problems were five problems (e.g., 5 × 94 = 470) and half were non-five problems (e.g., 8 × 63 = 504). False five problems violated arithmetic rules (i.e., the five rule, e.g., 5 × 17 = 87, the parity-rule, e.g., 86 × 5 = 435, both the five-and parity-rules, e.g., 5 × 42 = 411) or no arithmetic rules (e.g., 13 × 5 = 45). Problems were displayed superimposed on emotionally neutral or negative pictures. Behavioral results showed that negative emotions (a) did not affect participants' performance on true five and non-five problems, (b) influenced arithmetic performance on false five problems, and (c) impaired performance on problems that violated both the fiveand parity-rules but improved performance on false five problems violating no arithmetic rules. Electrophysiological data revealed that negative emotions led to (a) earlier P1 peak when participants verified true, non-five problems, (b) lower P300 and P600 amplitudes in central brain regions when participants verified false five problems that violated no-rule, (c) earlier N2 peak latencies in central brain regions and larger LPC amplitudes in right parietal regions while participants verified parity-rule violation problems, and (d) earlier N2 peak latencies in central brain regions and later N2 peak latencies in the right prefrontal brain regions while participants verified false, five problems violating both the five-and parity-rules. These findings demonstrate that negative emotions significantly alter key stages of arithmetic problem-solving by modulating neural activity related to encoding, detection of rule violations, and strategic execution, as evidenced by changes in the amplitude and latency of ERP components such as P1, N2, P300, P600, and LPC.