Simple arithmetic: evidence of an inhibitory mechanism to select arithmetic facts

The analysis of second trial depending on the type of first trial. It is important to note that the type of second trial (multiplication 2 vs. unrelated 2) could not be analyzed depending on the type of first trial (multiplication 1 vs. unrelated 1) due to a repetition effect that might have a different impact on the two conditions of trial 2. For example, while the solution 8 is repeated in the multiplication 2 condition: 2 + 6 = 8, after the multiplication 1 condition 2 + 4 = 8; the solution 10 is repeated in the unrelated 2 condition 4 + 6 = 10 after the unrelated 1 condition 2 + 4 = 10. Note, however, that this unbalanced repetition effect is avoided when multiplication 2 and unrelated 2 conditions are directly compared since in both conditions, half of the solutions were explicitly presented in the previous trial. Nevertheless, we performed additional analyses to evaluate the influence of Trial 1 on Trial 2 by avoiding the unbalanced repetition effect. Firstly, the Trial 1 (multiplication 1, unrelated 1 condition) x Trial 2 (multiplication 2, unrelated 2) interaction was significant, F(1,47) = 17.55, p < 0.05, η ² = 0.27. Afterward, we compared the two conditions that involved repetition: (a) multiplication 1–multiplication 2 condition vs. (b) unrelated 1–unrelated 2 condition. The RTs in Trial 2 were 100 ms slower in (a) vs. (b), F(1,47) = 6.18, p < 0.05, η ² = 0.12, suggesting that, in spite of the repetition effect, Trial 2 was difficult to perform when it included the result of multiplying the operands of Trial 1. However, the comparison between (c) multiplication 1–unrelated 2 vs. (d) unrelated 1 vs. related 2 was not significant, F(1,47) = 2.71, p > 0.05, η ² = 0.05 (57 ms difference). These two conditions did not involve repeating the result. Hence, the consequences of inhibiting the result of multiplying the operands of Trial 1 were only evident in Trial 2 when the result was visually present in both trials, the (a) condition.

Possible differences due to the gender of participants were examined. In Experiment 1, the results obtained in the first trial did not show differences between females and males, F(1,46) = 1.61, p > 0.05, η ² = 0.03, and the Gender × Type of addition result was not significant F < 1. In the second trial, Gender was not a significant variable, F(1,46) = 2.35, p > 0.05, η ² = 0.05, nor this variable interacted with type of results of previous trial, F < 1. In Experiment 2, the results of trial 1 did not show a significant effect of Gender, F < 1, and this variable did not interact with type of addition results, F < 1. In the second trial, Gender was not significant, F(1,31) = 1.12, p > 0.05, η ² = 0.03, and this variable did not interact with type of result of previous trial, F(1,31) = 1.20, p > 0.05, η ² = 0.04. Hence, there were no differences due to the gender of participants in this study.