The impact of visual sexual appeals on attention allocation within advertisements: an eye-tracking study

Abstract

To address a longstanding gap in the literature examining distraction effects resulting from the use of sexual appeals, a mixed-factors eye-tracking experiment (N = 108) was conducted to gauge how the use of sexual appeals in the form of partially clad, decorative female models impacts visual attention to competing ad components: model, product, advertising copy and the logo or brand. Findings suggest a visual distraction effect for ads containing models to operationalize sexual appeals. Although sexual appeals did not increase overall attention to the ad, gaze data reveal that for both men and women, visual attention directed specifically to models in ads with sexual appeals was greater than attention to models in non-sexual appeals. Moreover, this was more pronounced for men versus women. In addition, attention to ad copy and the brand suffered when sexual appeals were employed. These results provide greater specificity with respect to how sexual appeals impact consumer response and suggest that advertisers who wish to achieve strategic messaging goals through advertising copy should use visual sexual appeals with caution.

Keywords:

• Sexual appeals
• attention
• eye tracking

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 To ensure that no systematic bias was present across ad types in terms of the size of these elements, the area in pixels occupied by each element was calculated. A series of ANOVA tests where this measure served as the dependent variable verified no difference in each area of interest across ad types (all p > .05). Furthermore, number of words within the copy blocks for each category of ads was also compared across ad categories via ANOVA. That test likewise found no difference in the mean number of words within ads across categories.

2 Each ad was inserted after four pages of editorial content throughout the entire magazine. The placement of ads was first randomized and then counterbalanced by dividing the magazine into three parts, each containing one sexual appeal ad, one nonsexual appeal ad and one product-only ad.

3 Stimuli were presented on a 21 inch LCD monitor with 4:3 aspect ratio at 1280 × 1024 resolution. Gaze was tracked using an Applied Science Laboratories EyeTrac 6 control unit and remote desktop optics. The apparatus uses infrared light emitting diodes within the camera optics to corneal reflections, and the system continuously records the relative difference between this reflection and pupil center to track point of gaze. Gaze was sampled at 60 Hz, and data were recorded by Gazetracker software, which presented the stimuli and synchronized gaze data with stimuli.

4 The 15-min time limit is based on a pretest conducted in a computer lab with 21 identical personal computers in separated workstations. Twenty-one participants from the same population as the main study browsed the same magazine in a group session. Each magazine page was scanned and displayed in the middle of the computer monitor. Participants could press the spacebar on the keyboard to advance to the next page. On average, participants spent 12.41 min (SD = 7.67 min; range 3.18–34.28 min) to browse the entire magazine.

5 Data regarding attitude towards the ads and recall may be obtained from the first author. In general, results suggest that while males and females had similar attitudes towards the ads featuring nonsexual appeals and product-only ads, they differed in their attitude towards ads depicting sexual appeals. Females had significantly more negative attitudes towards these ads than males.

6 In the pretest, a sample of 30 participants (female n = 18; male n = 12) drawn from the same population as the main study evaluated all candidate ads. Presentation order of ads and dependent measures was randomized via presentation software. Descriptive statistics for ad originality, brand familiarity and ad familiarity were examined for all 27 candidate ads. No section of candidate ads effectively balanced across all three potential confounds while still maintaining factorial variation in rated sexuality. Therefore, nine were selected for use in the main study that achieved balance on two of the three potential confounds (i.e. ad and brand familiarity).

7 For all multiple comparisons between three categories of ad type, Bonferroni’s correction for familywise error was employed.

Additional information

Notes on contributors

R. Glenn Cummins

R. Glenn Cummins (Ph.D., University of Alabama) is an associate professor of journalism and creative media industries at Texas Tech University. His research focuses on how production elements interact with content properties and individual characteristics to influence message processing and response.

Zijian Harrison Gong

Zijian Harrison Gong (Ph.D., Texas Tech University) is an assistant professor in the Department of Advertising at Texas Tech University. His research focuses on motivated processing of media to understand how message execution elements such as design and layout impact selective attention, recall, and attitude formation, and how source and message factors impact the persuasiveness of media content.

Tom Reichert

Tom Reichert (Ph.D., University of Arizona) is professor and dean of the College of Information and Communications at the University of South Carolina. His research and teaching focus on advertising and media content and effects, including issues and concerns regarding the uses and effects of advertising on professional practice and culture.