Dissecting the Reaction Times of Global and Local Processing

Reaction time studies using hierarchical visual stimuli typically demonstrate that global processing is faster than local processing (global precedence effect), and that global interference during a local task has a greater impact than the opposite (global interference effect). Previous examinations of individual differences in these reaction times have sometimes provided evidence that these effects are modulated by various clinical conditions (e.g., autism), but these reports are inconsistent in the literature. This may be because reaction times are a blunt measure of processing speed, since they provide only a sum of the durations of the many different processes that occur between stimulus and response (including afferent transmission, decision, and efferent transmission), and they are prone to distortion due to speed-accuracy tradeoffs. Here, we used a compelled response task (Stanford et al., 2010) to tease apart the duration of decision processes from that of non-decision processes (the sum of afferent and efferent delays). Participants viewed hierarchical C’s (large C’s made of small c’s, both of which could be open to the right or left in a mirror-reversed fashion) and reported their orientations at either the global or local level (in a blocked design) with right or left button presses. We found that although the global interference effect was reflected in almost equal measures in non-decision and decision times, the global precedence effect was mostly comprised of an increased duration for local decisions compared to global decisions with very little difference in the durations of non-decision processes. The ability of the compelled response task to provide more precise measures of the durations of the underlying components of a response indicates benefits of its use over typical reaction time measures for examining individual differences in sensorimotor processing.