Public Estimates of Cancer Frequency: Cancer Incidence Perceptions Mirror Distorted Media Depictions

Abstract

Compared with incidence rates, certain cancers are over- or underrepresented in news coverage. Past content analytic research has consistently documented these news distortions, but no study has examined whether they are related to public perception of cancer incidence. Adults (N = 400) completed a survey with questions about perceived cancer incidence, news consumption, and attention to health news. Cancer incidence perceptions paralleled previously documented news distortions. Overrepresented cancers were overestimated (e.g., blood, head/brain) and underrepresented cancers were underestimated (e.g., male reproductive, lymphatic, thyroid, and bladder). Self-reported news consumption was related to perceptual distortions such that heavier consumers were more likely to demonstrate distorted perceptions of four cancers (bladder, blood, breast, and kidney). Distortions in risk perception and news coverage also mirrored discrepancies in federal funding for cancer research. Health care professionals, journalists, and the public should be educated about these distortions to reduce or mitigate potential negative effects on health behavior and decision making.

Notes

Note. Top five cancers depicted in mainstream/general audience newspapers by content analysis. Cancers are listed horizontally from most to least coverage. Freimuth et al. (1980) did not provide a fourth or fifth most reported cancer.

Note. Perceived incidence rank is represented in two columns. Average perceived incidence rank is the mean rank (and standard deviation) for each cancer type. Absolute perceived incidence rank translates the mean rank into an absolute ranking. For example, breast cancer had a mean ranking of 2.93, which translates to 1st in perceived incidence rank whereas bladder cancer had a mean ranking of 10.62, which translates to 15th in perceived incidence rank. Actual incidence rank was calculated using 2007 incidence rate data from the National Program of Cancer Registries. Perceptual distortion was calculated by subtracting absolute perceived incidence rank from actual incidence rank. A positive score means the incidence of the cancer was overestimated and a negative score means it was underestimated. News difference scores were culled from Jensen et al. (2010).

Note. Ordinal regression analysis predicting average perceived incidence rank with demographics and news consumption. Coefficients with standard errors presented in columns. Perceived incidence rankings were reverse coded to ease interpretation for this analysis. For news consumption, a positive coefficient means with increased consumption a cancer was perceived to be more frequent. A negative coefficient means with increased consumption a cancer was perceived to be less frequent. Race was coded 0 = not Caucasian, 1 = Caucasian. Gender was coded 0 = male, 1 = female. Age, education, and news consumption are all continuous variables were higher scores equate to older, more educated, and heavier consumption. All significance tests were two-tailed. For each χ ², df = 5.

†p < .10. *p < .05. **p < . 01.