Most research on images in surveys has examined how images affect answers to questions, which are known as visual context effects. Less frequently studied is how respondents process images in surveys and how variation in processing impacts observing visual context effects. Using data from two eye-tracking studies, we found that less than half of the respondents looked at the images in our surveys and that rates of looking at the images and time spent processing them significantly differed by gender. Additionally, we found that those who looked at the images for at least 60 milliseconds had their responses influenced by the images. The visual context effects did not occur for respondents who did not fixate on the images. Our findings suggest researchers must be cautious when placing images in surveys, given the possibility of visual context effects and variation of these effects by whether respondents fixate on the images, which could vary among subgroups.