Advancing the Extended Parallel Process Model Through the Inclusion of Response Cost Measures

Abstract

This study advances the Extended Parallel Process Model through the inclusion of response cost measures, which are drawbacks associated with a proposed response to a health threat. A sample of 502 college students completed a questionnaire on perceptions regarding sexually transmitted infections and condom use after reading information from the Centers for Disease Control and Prevention on the health risks of sexually transmitted infections and the utility of latex condoms in preventing sexually transmitted infection transmission. The questionnaire included standard Extended Parallel Process Model assessments of perceived threat and efficacy, as well as questions pertaining to response costs associated with condom use. Results from hierarchical ordinary least squares regression demonstrated how the addition of response cost measures improved the predictive power of the Extended Parallel Process Model, supporting the inclusion of this variable in the model.

Notes

¹We use sexually transmitted infection (STI) in the body of this article but used sexually transmitted diseases (STDs) in the study materials. STI is the term more frequently used by clinicians and health promotion agencies such as the Center for Disease Control and Prevention, but informal consensus from our students was that STD was their common vernacular and that STI may even cause confusion for some of them.

Note. All scales were measured with 6-point Likert scales ranging from 1 (strongly disagree) to 6 (strongly agree).

²Besides the large percentage of Caucasian respondents, there were also 6% African Americans, 2.6% Hispanics, 10.9% Asian or Pacific Islanders, 1.1% multi-racial, and 1.3% other racial groups in our sample. We did not measure respondents' current relationship status and their sexual identities as a result of privacy concerns for research that involves human subjects.

³The EPPM argues that threat appraisal and efficacy appraisal work together to determine whether danger-control process or fear-control process ensues. When it is impossible to explore this mechanism through artificially manipulated high/low conditions, such as with the cross-sectional data in our study, we explored the interaction between these variables, hoping to identify the key variables at work that lead to stronger danger-control behavioral intention.

Note. Standardized regression coefficients (betas) except where indicated. All variables in the same block were entered together (listwise deletion). Final model included the only significant interaction term.

*p < .10. **p < .05. ***p < .01.

Note. Standardized regression coefficients (betas) except where indicated. All variables in the same block were entered together (listwise deletion). Final model included the only significant interaction term.

*p < .10. **p < .05. ***p < .01.

⁴Three of the response costs measures (mood, value, and relational costs) are related to self-efficacy (r ranges from −.13 to −.32, p < .01). However, response efficacy and self-efficacy are also moderately correlated (r = .25, p < .001). These moderate correlations (Cohen et al., 2003) are unsurprising, as these three variables are all part of the Protection Motivation Theory's coping appraisal process.