Abstract
Measures of dynamic visual processing (e.g. motion processing) have received increasing interest as potential markers of subtle neural dysfunction in clinical conditions as varied as dyslexia and schizophrenia. However, there has been little work establishing the basic psychometric properties of such tasks. This research investigates the reliability and validity of a number of dynamic visual processing tasks in a naive, non-clinical sample in order to identify those that may show potential for use in clinical samples. The particular hypotheses that are tested for each task are a) that it is capable of efficiently isolating reliable individual differences, and b) that these individual differences show convergent validity on theoretically related tasks and discriminant validity on theoretically unrelated tasks. Additionally, estimates are made of the extent to which reliability of performance drops over the course of a 1-2 hour experimental session. Such establishment of reliability and validity of individual differences on a task, as well as robustness to fatigue, are essential prerequisites for using a task as a clinical instrument.