The Effects of Heuristic Cues, Motivation, and Ability on Systematic Processing of Information About Breast Cancer Environmental Factors

Abstract

The heuristic systematic model is used to investigate how ability, motivation, and heuristic message cues predict knowledge scores for individuals receiving messages written for different literacy levels about 3 environmental risk factors for breast cancer. The 3 risk factors were the roles of genetics, progesterone, and ingesting perfluorooctanoic acid in breast cancer risk. In this study, more than 4,000 women participated in an online survey. The results showed support for the hypotheses that ability (measured as education, number of science courses, and confidence in scientific ability) predict knowledge gain and that those individuals who presented with the lower literacy level message had significantly higher knowledge scores across all 3 message topics. There was little support for motivation or heuristic cues as direct predictors of knowledge gain across the 3 message topics, although they served as moderators for the perfluorooctanoic acid topic. The authors provide implications for health communication practitioners.

Acknowledgments

This research was supported by the Breast Cancer and the Environment Research Centers grant number U01 ES012800 from the National Institute of Environmental Health Sciences, and the National Cancer Institute, National Institutes of Health, Department of Health and Human Services. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Institute of Environmental Health Sciences or of the National Cancer Institute/National Institutes of Health.

Notes

¹Gunning Fog and Fry readability level tests were also conducted on the lower literacy level messages. The genetic, perfluorooctanoic acid, and progesterone messages received a 9.1 and 7.7, 9.3 and 6.5, and 9.14 and 10.3 on these instruments, respectively. To see the messages, contact the corresponding author.

²Contact the corresponding author for the results of a manipulation check on perceived message difficulty.

³We also ran regressions with the demographic variables of age and ethnicity in the first block. The first block of each of those regressions was nonsignificant. Please write to the corresponding author if you wish to obtain the results of those regressions.

Note. ΔF(11, 979) = 12.451, p < .001. ΔR ² = 0.123, adjusted R ² = 0.113.

*p < .05. **p < .01. ***p < .001.

Note. Only significant interaction terms are included in the tables. Please contact the corresponding author for the other interaction terms if you wish to see them.

With regard to multicollinearity, each of the first-order predictors had a variance inflation factor of less than 2.0. As a rule of thumb, Cohen, Cohen, Aiken, and West (2002) noted that a variance inflation factor of 10 or greater is considered to be an indicator of problematic multicollinearity. However, when the product terms were added, only two variables in the model had a variance inflation factor greater than 10: the high school education dummy variable and the product term of high school graduate education level and number of scientific courses.

For the first block, ΔF(11, 1124) = 26.448, p < .001, R ² = 0.206, adjusted R ² = 0.198. For the second block, ΔF(48, 1076) = 1.426, p = .032, ΔR ² = 0.048, adjusted R ² = 0.212.

*p < .05. **p < .01. ***p < .001.

Note. For first block, F(12, 1168) = 12.219, p < .001, R ² = .112, adjusted R ² = 0.102.

*p < .05. **p < .01, ***p < .001.