Tobacco Counteradvertisements Aimed at Bicultural Mexican American Youth: The Impact of Language and Theme

Abstract

The growing population of Mexican American youth and the increasing smoking rates in this population present a considerable public health challenge. Tobacco counteradvertisements have demonstrated their ability to shape attitudes, behavior, and public policy, but little is known about the most effective ways to adapt messages aimed at this audience. To explore key variables that can affect success, a study was conducted with 249 Mexican American middle-school youth from a U.S./Mexico border community to examine the effectiveness of language (English, Spanish, or a combination of English and Spanish) and theme (secondhand smoke, antitobacco social norms, and tobacco industry manipulation) in print tobacco counteradvertisements. Measures included ad preferences, acculturation, and tobacco-related attitudes and behavior. Results showed that although a large percentage identified with the Mexican American rather than the Anglo American culture and spoke Spanish in selected contexts, readability was greater for ads in English, and participants rated the English ads as most effective. The social norms counteradvertisement was preferred overall. Potential implications for health communication and policy are discussed.

Acknowledgments

This study was conducted as part of a larger research project funded by the National Institute on Drug Abuse P50 DA007074.

Notes

¹For k = 1 and 2, γ_k3 and γ_k4 are set a priori equal to 0 because no interaction effect between theme and Spanish/English communication is expected. Likewise, γ₃₅ and γ₄₅ are set a priori equal to 0, and, in the model for readability only, γ₀₅, γ₁₅, and γ₂₅ are set a priori to 0.

²To obtain standardized coefficients in a multilevel model, all independent variables must first be standardized. We believe the fixed coefficients are more interpretable if the independent variables are not standardized.

*p < .05; **p < .01.

³Each culture identification measure was divided into 3 levels—high, medium, and low—according to the procedure outlined by Oetting and Beauvais (1991). Low was defined as an average score less than 2; a medium score was from 2 to 3; and a high score was greater than 3. Overall, 41% of respondents fell into the high level of Mexican American identification and the medium level of White American identification. The next highest percentage, 30%, was for the combination of high Mexican American identification and low White American identification. Only 9% of respondents considered themselves high in both Mexican American and White American identification.

*p < .05; **p < .01; ^a p < .10.