Young Adults' Responses to News About Sunscreen and Skin Cancer: The Role of Framing and Social Comparison

Abstract

Young adults read a news article about skin cancer that used a gain frame (e.g., using sunscreen keeps skin healthy) or a loss frame (e.g., not using sunscreen increases skin cancer risk), and included a personal exemplar (healthy or ill). Compared to a control group, both frames increased intentions to use sunscreen. Planned SPF was higher for men in the loss frame but was unaffected by framing for women. The framing manipulation had opposite effects on sun protection behaviors among individuals who were high versus low in perceived similarity to the exemplar. Overall, the gain frame was more effective for people who were high in similarity, whereas the loss frame was more effective for those low in similarity. In addition, individuals who felt more similar to the exemplar reported more personal risk in the loss frame but stronger intentions to use sun protection behaviors in the gain frame. Last, social comparison orientation was associated with more personal risk in the gain frame (suggesting upward comparison) but lower intentions to engage in sun protection behaviors regardless of the frame. Interpretations of the findings and implications for the use of exemplars in health messages are discussed.

Notes

¹The ethnicity of the individual in the example was not specified or made apparent. Most print news articles do not specify the race or ethnic background of interviewees (unless a photo is used), whereas gender is commonly identified through the use of names and pronouns. Moreover, perceived similarity in skin tone or susceptibility to sunburn would probably be more important than ethnic similarity per se, and this would be very difficult to manipulate effectively.

²A separate item asked respondents to estimate the probability that they will develop skin cancer during their lifetime, using a response between 0% chance (no possibility) and 100% chance (definitely will develop) (M = 33.32%, SD = 23.45%). The analyses of this item yielded results that were very similar to those obtained using the rating scale of personal risk. Thus, for the sake of simplicity, this item was eliminated from the results.

³When the gender by frame interaction was included in the regression analyses after framing, the other regression results remained the same. Thus, to avoid redundancy with the analyses of covariance, the interaction was not included in the reported regression analyses.

⁴Follow-up analyses examined whether gender and perceived similarity to the exemplar were related to involvement with the article (which did not differ for the two frames). The analyses confirmed that involvement was higher among women (M = 4.50) than men (M = 3.98), t(146) = 2.05, p < .05, and that perceived similarity was associated with greater involvement, r = .28, p < .001.