How do we estimate magnitudes? Validating the magnitude stroop task

Humans have the ability to make spontaneous and rapid estimates of the approximate number of items they see, without explicitly counting them (e.g., estimating the number of fruits on a tree). However, most of our decisions are based on comparing symbolically represented numbers, such as Arabic numerals. Here, we validate a variant of the Stroop task, the "Magnitude Stroop," which investigates differences in behaviour when numbers are represented as magnitudes versus symbols. In this task, stimuli are presented simultaneously as numerical symbols and magnitudes (e.g., the number 1 is represented by vertically aligned dots). In two studies (online N = 204 and offline N = 60), participants were asked either to select the stimulus with more dots or with the largest number. These two rules are occasionally equivalent but at times not, hence creating a possible Stroop interference effect. We find that the Stroop effect is stronger when participants are asked to select the stimulus with more dots and ignore the largest number, compared to when they should ignore the number of dots and select the largest number. These findings suggest that more research on how people estimate magnitudes across modalities is needed and we validate a method which could help investigations in this topic.