Money counts: effects of monetary vs. purely numerical values on the mental representation of quantities

It has been established that humans use different cognitive models to represent and process numerical quantities. In this study, we investigated whether the representation of monetary values fundamentally differs from the representation of numbers. We also examined the influence of both socio/economic factors and mathematical ability on such representation. A group of adults (N = 272) were tested anonymously with a variant of the number-to-position task (Siegler & Opfer, 2003). They were asked to position on a horizontal line quantities expressed either in numerical format (e.g., 50) in the "Number" conditions or as monetary values (e.g., 50) in the "Money" conditions. The extremes of the line consisted either of specific values (i.e. "2 or 2" and "503 or 503") in the "Fixed" conditions or of unspecific concepts of quantity (e.g., "little" and "a lot") in the "Fuzzy" conditions. A linear model, as opposed to a logarithmic one, provided the best fit of group average data in all experimental conditions except for the "Money-Fuzzy" condition. The percentages of absolute error were significantly larger for Money stimuli than Number stimuli in both Fixed and Fuzzy conditions. This is consistent with the law of diminishing marginal utility, which entails that the value of monetary quantities is described by a concave curve rather than a linear relationship. As expected from previous research, participants who were more used to spending large quantities of money were closer to the linear representation model. Participants with higher mathematical abilities represented numerical values more closely to a linear model, but no such effect was found for monetary quantities.