Abstract
With STEM education currently an international priority, efforts to isolate math-related aspects of human cognitive variation are of high interest. Here, we marshall the recruiting power of our citizen science project TestMyBrain.org to ask whether the classic experimental paradigm of multiple object tracking (MOT) can be used to capture math-related mechanisms of cognitive variation. We developed a brief (8-10 min), sensitive (Cronbach's alpha = .88), web-based MOT test and used it, along with other reliable (alpha > .80), well-validated tests to assess 19,724 participants (10,014 male) of varying ages (5th, 50th, and 95th percentiles 14, 25, and 58 years, respectively), educations (e.g. 34% of 25+ year olds had no more than a high school diploma), and ethnicities (46% non-Caucasian). Nearly two orders of magnitude larger than the entire prior literature on individual differences in MOT, this study represented an unprecedented opportunity to better understand how and why MOT performance differs across individuals. Four major findings emerged. First, MOT dissociated strongly from sustained visual attention (r(1930)=-.03) and vocabulary (r(2810)=-.09), indicating a remarkable cognitive specificity. Second, MOT associated strongly with spatial working memory (r(10841)=.44) and rapidity of spatial attentional switching (r(533)=.41), documenting MOT's validity as a core measure of visuospatial processing. Third, MOT robustly predicted SAT-math scores (r(2467)=.29), yet far less so SAT-verbal scores (r(2467)=.09), revealing a strong, specific connection to math potential. Fourth, STEM-majors scored substantially (up to 10-20 percentile points) higher on MOT than non-STEM-majors. Taken together, these findings indicate that our brief, web-based measure of MOT performance captures a core, math- and STEM-related aspect of visuospatial ability. We suggest that tests of MOT be considered a relevant, feasible addition to future studies of number cognition, quantitative abilities, and math/STEM education.
Meeting abstract presented at VSS 2016