The Visuomotor Mental Rotation Task: Visuomotor Transformation Times Are Reduced for Small and Perceptually Familiar Angles

ABSTRACT

In the visuomotor mental rotation (VMR) task, participants point to a location that deviates from a visual cue by a predetermined angle. This task elicits longer reaction times (RT) relative to tasks wherein the visual cue is spatially compatible with the movement goal. The authors previously reported that visuomotor transformations are faster and more efficient when VMR responses elicit a degree of dimensional overlap (i.e., 0° and 5°) or when the transformation involves a perceptually familiar angle (i.e., 90° or 180°; K. A. Neely & M. Heath, 2010b). One caveat to this finding is that standard and VMR responses were completed in separate blocks of trials. Thus, between-task differences not only reflect the temporal demands of the visuomotor transformations, but also reflect the temporal cost of response inhibition. The goal of this study was to isolate the time cost of visuomotor transformations in the VMR task. The results demonstrated that visuomotor transformations are more efficient and effective when the response entails a degree of dimensional overlap between target and response (i.e., when the angular disparity between the responses is small) or when the transformation angle is perceptually familiar.

Keywords:

• rapid aiming movements
• response-planning processes
• stimulus-response mapping
• visuomotor processing

ACKNOWLEDGMENTS

A Discovery Grant from the Natural Sciences and Engineering Research Council of Canada and a Major Academic Fund from the University of Western Ontario supported this research.

Notes

1. Our use of the term dimensional overlap is consistent with the stimulus-response compatibility literature describing degrees of commonality or similarity between a stimulus and a response (Kornblum, Hasbroucq, & Osman, 1990). In terms of the present work, such commonality is determined by the spatial disparity between the stimulus-driven and the VMR responses.

2. Although some VMR studies have examined eye movements (de’Sperati, 1999; Fischer, Deubel, Wohlschlager, & Schneider, 1999), our discussion of the VMR literature is restricted to limb movements.

3. There is no evidence from Georgopoulos and Massey (1987) to suggest the mental rotation model (MRM) predicts RT only for the set of angles used in their manuscript. Moreover, it is important to note that exemplar neurophysiological evidence describes mental rotation of the neuronal population vector for 90° (see Figure 1 of Georgopoulos, Lurito, Petrides, Schwartz, & Massey, 1989).

4. On average, participants reported one directional error per block of trials. Some participants completed several blocks in a row without making an error. To determine (offline) if participants initially specified their reach trajectory in line with the task instructions, we calculated mean direction at peak acceleration. The results of this analysis were consistent with the results for mean direction at ultimate movement endpoint. For the standard trials, there was no effect of instruction angle, F(6, 60) = 1.07, p = .392. Specifically, mean direction was in line with the veridical target coordinates. For the VMR trials, the F statistic demonstrated an effect of instruction angle, F(6, 60) = 257.93, p < .001, such that the direction of the reach trajectory increased as a function of increasing instruction angle. Post hoc analysis demonstrated that mean direction was different for each instruction angle (all ps < .003).

5. To determine whether RT was normally distributed, we examined the ratio of the skewness value to its associated standard error. The skewness statistic was 17.73 and thus we elected to evaluate the medians.

6. The reference in the preceding cited literature focus on spatial data, whereas the examples cited focus on temporal data.