Working memory load as a novel tool for evaluating visual analytics

The current visual analytics literature highlights design and evaluation processes that are highly variable and situation dependent, which raises at least two broad challenges. First, lack of a standardized evaluation criterion leads to costly re-designs for each task and specific user community. Second, this inadequacy in criterion validation raises significant uncertainty regarding visualization outputs and their related decisions, which may be especially troubling in high consequence environments like those of the intelligence community. As an attempt to standardize the ldquoapples and orangesrdquo of the extant situation, we propose the creation of standardized evaluation tools using general principles of human cognition. Theoretically, visual analytics enables the user to see information in a way that should attenuate the user´s memory load and increase the user´s task-available cognitive resources. By using general cognitive abilities like available working memory resources as our dependent measures, we propose to develop standardized evaluative capabilities that can be generalized across contexts, tasks, and user communities.