Funded by the Annenberg Public Policy Center, EPIC Investigators developed a tobacco use survey for young adults. Using data from this tobacco use survey, researchers with the Theory & Methods Core are exploring the relationships between intentions to (a) smoke in the next month, (b) try to quit in the next 3 months and (c) quit in the next 3 months and reported smoking behaviors (including 3 month quitting attempts and quitting) obtained 12 months later. All respondents were current smokers at baseline. Although intentions are quite good predictors of quitting attempts and quitting behavior, there are a number of respondents who are either unable or unwilling to act on their intentions. Interestingly, in addition to those who intend to try to and/or who intend to quit but do not do so, there are a substantial number of smokers who say they definitely will not try to quit and actually do voluntarily stop smoking for 24 hours or more. Researchers at the EPIC Center are investigating how underlying beliefs, attitudes, perceived norms and feelings of self-efficacy are related to both success and failure in acting upon one's intentions. Preliminary data analyses indicate that these traditional Integrative Model variables contribute very little to our understanding of why people do or do not act on their intentions. Researchers are now looking at other variables that may help explain the intention-behavior gap.

It has been suggested that the construct of "willingness" can account for variance in behavior that is not captured by measures of intention. Again using data from the tobacco use survey, researchers working with the Theory & Methods Core are investigating the role of "willingness" as a determinant of smoking, trying to quit and actual quitting. Preliminary findings suggest that, as expected by the prototype-willingness model, willingness does explain some variance in these behaviors over and above the variance accounted for by intention. However, it appears that this additional variance is due to the fact that the willingness construct taps self-efficacy as well as intention. That is, according to the theory of planned behavior and the integrative model that serves as a framework for the EPIC Center's research, both intentions and perceived self-efficacy (or perceived behavioral control) are viewed as determinants of behavior. When measures of both intention and self-efficacy are included in the predictive model, "willingness" no longer contributes to the prediction of behavior. Researchers are currently trying to determine whether efficacy has a direct effect on behavior or if it serves to moderate the intention-behavior relationship. A poster describing this project’s initial findings was presented at the annual meeting of the Society for Research on Nicotine and Tobacco in March 2005.

Social scientists and lay theorists agree that social norms influence behavior. Studies using TRA/TPB generally report significant correlations between subjective, or injunctive norms (what others think I should do) and behavioral intentions. Yet, after accounting for the effects of attitudes and perceived behavioral control, injunctive norms often fail to account for a significant portion of the variance in behavioral intentions. Some evidence suggests that including measures of descriptive norms (what others are doing) explains a significant amount of variance over and above models using attitudes, behavioral control, and injunctive norms as predictors. What is less clear are the relations among injunctive and descriptive norms and behavioral intentions. Some studies have combined measures of injunctive and descriptive norms into a single measure of normative pressure, and others have use structural equation models to examine whether injunctive and descriptive norms are lower-order factors that comprise a higher-order normative pressure construct. Many individual studies that examine these issues use a select segment of the population (e.g., adolescents, adults over 40) and most examine only a small set of behaviors and are thereby unable to answer in a global sense how and when norms influence behavior.

We are currently addressing these and related questions from three angles.

1. A large scale meta-analysis of the TRA/TPB literature designed to examine the conditions (e.g., populations, behaviors, types of norms, and item wording and scoring) that moderate the effect of normative influences on health-related behaviors.
2. Dedicated experiments addressing specific measurement, scoring, and modeling questions across numerous behaviors and populations.
3. Continued analysis of existing datasets and inclusion of items on other CECCR-related survey instruments.

This research, led by Barbara Kahn, PhD, focuses on the effects of generated anxiety on subsequent screening test adherence. Researchers suggest that the hospital waiting room may prime previous testing anxiety that can lead to delay in subsequent testing intentions. This research will investigate the effects of various other kinds of priming manipulations to determine how they might affect testing and general health anxiety. Researchers will explore primes that are promotion-focused, focusing individuals on their hopes, aspirations and nurturance needs versus prevention-focused where people are focused on their duties, responsibilities and security needs. We hypothesize that these different types of priming will affect patients testing anxiety differentially as a function of their false positive history. For patients who have false positive histories, a prevention-focused prime will increase subsequent testing adherence as compared to a promotion-focused prime. For patients with no false positive histories, but rather normal previous screening results, a promotion-focused prime will increase subsequent testing adherence.

This project, led by Oscar Gandy, PhD, focuses on the nature of risk perception among African Americans. The literature on optimistic bias is extensive, but there is very little about optimistic bias among African Americans, and even less that examines the relationship between optimistic bias and health related screening and prevention activities. This project will integrate the literature on health risk perception and related literatures on optimistic bias, third-person effect and racial identity among African Americans. It is identifying relevant data sets, such as the MIDUS [Midlife development in United States] dataset and those made available through the Kaiser Network, that combine health locus of control (HLOC), risk assessment, and cancer screening and prevention. While there is some work being done to explore HLOC among African Americans, including some innovative investigations of religious subscales, there is a need for an integrative analysis of these risk-related literatures that takes African American cultural perspectives into account. Preliminary results suggest that belief in the influence of God may be a factor in HLOC, especially among African Americans. Preliminary analyses of the data collected for Hughes-Halbert's pilot project has examined the ways in which HLOC relates to personal health experience and orientations toward genetic testing. Future analyses will explore how religiosity affects those relationships.