Project Overview +

This study explores the effects images have on the amount of time spent reading text that is shown alongside images. Specifically, the project examines the time a smoker spends reading 3 brief testimonials, as well as the time spent looking at the images shown alongside the testimonials.

Aims +

To see whether different types of images paired with a testimonial story matter to testimonial readers. Do they look at the images differently or not? We will test 3 different images:

  • matched person image (matched to race, age, gender)
  • mismatched person image
  • a standard health promotion image

Participants +

85 individuals ages 21-35 and 50-65, who smoke at least 5 cigarettes a day, have smoked at least 100 cigarettes in their lifetime, have normal or corrected-to-normal vision, and are interested in quitting smoking. Participants recruited through local newspaper ads, flyers, and Engage (the University of Michigan's clinical research study volunteer website).

Intervention +

Participants read three testimonials from three different people who successfully quit smoking. Each of the three testimonials will include one image. Images are randomly assigned to each of the three testimonials such that the participant will see one person-matched image (age, gender, race), one person-mismatched image, and a sunset image alongside a testimonial. Participants wear an eye-tracking measurement device to capture their eye movements on the computer monitor as they view each testimonial/image combination.

Findings +

We collected eye movements as individuals read health promotion narratives paired with matched or mismatched faces, and assessed the persuasiveness of the narratives.

Eighty-five smokers each read the same 3 narratives from "successful quitters". Each narrative was paired in a counterbalanced order with either a face that matched the smoker in age, race, and gender, a face that mismatched the smoker on the three variables, or a neutral picture.

Eye movements were analyzed in terms of two critical regions: the image and the text. Overall, the narratives with in-group narrators were rated most persuasive, χ2 (2, N = 84) = 84.0, p < .001. While viewing the narratives, participants fixated the matched face more than the mismatched and neutral images, though time on text did not differ, F(1,84) = 5.74, p = .02.

In addition, people made more saccades between the text and the image when the image was a matched picture, F(1,84) = 5.74, p = .02. Of those that selected the narrative with a narrator’s image (matched or mismatched) as most persuasive, participants made more fixations, contrast F(1,59)= 6.13, p = .01, and have longer dwell times, contrast F(1,59)= 6.83, p = .01, on the corresponding image than others.

However, the minority of participants who chose the narrative with a neutral picture as most persuasive spent relatively more time on the chosen narrative than others and no differences in time on the pictures , contrast F(1, 23) = 3.12, p = .09, and contrast F(1, 23) = 3.36, p = .08, respectively.

These results extend prior evidence of an in-group persuasion advantage to the field of health communication and reveal potential individual differences in how health promotion narratives are consumed.

Conclusion +

The eye tracking data supported people's subjective reports of which testimonial convinced them more. The results of the current study argue for the use of matched pictures, in terms of age, race, and gender, when giving people testimonials as part of a smoking cessation intervention.

Eye Tracking Tailored Photos

09/01/2007 - 08/31/2008

Sponsor(s)

National Cancer Institute

Principal Investigator:

Hannah Faye C. Chua, PhD