A meta-analysis on the effect of expertise on eye movements during music reading

The current meta-analysis was conducted on 12 studies comparing the eye movements of
expert versus non-expert musicians and attempted to determine which eye movement
measures are expertise dependent during music reading. The total dataset of 61 comparisons
was divided into four subsets, each concerning one eye-movement variable (i.e., fixation
duration, number of fixations, saccade amplitude, and gaze duration). We used a variance
estimation method to aggregate the effect sizes. The results support the robust finding of
reduced fixation duration in expert musicians (Subset 1, g = -0.72). Due to low statistical
power because of limited effect sizes, the results on the number of fixations, saccade
amplitude, and gaze duration were not reliable. We conducted meta-regression analyses to
determine potential moderators of the effect of expertise on eye movements (i.e., definition
of experimental groups, type of musical task performed, type of musical material used or
tempo control). Moderator analyses did not yield any reliable results. The need for
consistency in the experimental methodology is discussed.