Arithmetic is not arithmetic: Paradigm matters for arithmetic effects.

Research on arithmetic uses different experimental paradigms. So far, it is unclear whether these different paradigms lead to the same effects or comparable effect sizes. Therefore, this study explores how different experimental paradigms influence mental arithmetic performance, focusing on understanding the potential differences and similarities in cognitive processes between paradigms. Six paradigms were systematically compared: decision paradigms (verification, forced-choice, delayed forced-choice) and production paradigms (written production, verbal-keyboard production, and simple verbal production). The results show consistent arithmetic effects related to operation (addition vs. subtraction) and task difficulty (with or without carry/borrow) across all paradigms, particularly in reaction time measures. However, accuracy varied between paradigms, with verbal-keyboard production and simple verbal production paradigms showing higher effect sizes for accuracy measures. These findings underscore the importance of considering each paradigm's specific demands and characteristics in arithmetic research, suggesting that paradigm selection can influence the observed outcomes. Our study provides critical methodological insights that can guide future research in the design and interpretation of arithmetic tasks, enhancing the reliability and ecological validity of findings in numerical cognition.