How Images Draw the Eye: An Eye-Tracking Study of Composition

Abstract

In his instructional art book, Andrew Loomis provides images and corresponding diagrams that indicate how the composition of the image should guide the viewer's eye. Using these images, we examined whether participants would follow the suggested cues. Participants' eyes were tracked as they viewed the images, allowing us to take measures of where they entered and exited the image, whether they attended to the focal part of the image, and what path they followed between these components. These measures could then be compared with Loomis' suggestions, to determine if the elements did indeed have the proposed influence. While viewers were attracted to the focal points, and spent the most time examining these, they did not use the entry and exit points marked by Loomis, and the suggested viewing paths were not closely followed. It appears that Loomis' suggested elements of composition do not strongly influence viewers' eye movements.

Does pictorial composition guide the eye? Investigating four centuries of last supper pictures

“…Our study thus provides strong evidence for the traditional assumption in art literature that pictorial composition guides the movement of the eye. We thus contradict the results of Garbutt and Spehar (2014) and Kirtley (2018) who denied such correlations. However, this is not a real contradiction.…”

The role that composition plays in determining how a viewer looks at landscape art

“…This strengthens prior findings ( 19 ) that composition does influence the scanpath, but does not dictate the scanpath of the viewer. The results are similar to those of Kirtley ( 29 ), confirming that focal points attract and keep attention but that the entry and exit points are largely not used as intended. Furthermore, it confirms that the compositional line guided the eye but did not dictate either order or direction.…”

“…Results indicated that focal points were successful in attracting and keeping attention, but the intended entry and exit points were not used. Furthermore, the compositional line was not closely followed and the compositional elements did not exert a significant influence over eye movements (Kirtley, 2018). Interestingly, children rely on bottom-up processing when free-viewing a painting, but revert to top-down processing when they view the painting again after being given background information.…”

“…In this manner, 35 paintings with known construction of composition (as explained by the artists themselves in their instructional texts) were chosen for inclusion in the study. This method of stimuli inclusion is similar to the Kirtley ( 29 ) study, where an instructional text was also used to identify paintings with a known compositional line. Therefore, a clear intended scanpath (as envisaged by the artist’s composition) could be extrapolated for each painting.…”

“…Results indicated that focal points were successful in attracting and keeping attention, but the intended entry and exit points were not used. Furthermore, the compositional line was not closely followed and the compositional elements did not exert a significant influence over eye movements ( 29 ).…”