Dot comparison stimuli are not all alike: the effect of different visual controls on ANS measurement

The most common method of indexing Approximate Number System (ANS) acuity is to use a nonsymbolic dot comparison task. Currently there is no standard protocol for creating the dot array stimuli and it is unclear whether tasks that control for different visual cues, such as cumulative surface area and convex hull size, measure the same cognitive constructs. Here we investigated how the accuracy and reliability of magnitude judgements is influenced by visual controls through a comparison of performance on dot comparison trials created with two standard methods: the Panamath program and Gebuis & Reynvoet's script. Fifty-one adult participants completed blocks of trials employing images constructed using the two protocols twice to obtain a measure of immediate test-retest reliability. We found no significant correlation between participants' accuracy scores on trials created with the two protocols, suggesting that tasks employing these protocols may measure different cognitive constructs. Additionally, there were significant differences in the test–retest reliabilities for trials created with each protocol. Finally, strong congruency effects for convex hull size were found for both sets of protocol trials, which provides some clarification for conflicting results in the literature.