Presented at the 2014 AAPOR conference, PRRI CEO Robert P. Jones, PRRI Research Director Daniel Cox, and Juhem Navarro’s conference paper explores how social desirability bias impacts self-reported religious attendance, affiliation, salience and belief. The researchers compare responses to identical questions on two surveys: one administered by live interviewers on the telephone and one self-administered online. The authors find that respondents who took the survey via telephone were more susceptible to social desirability bias; they were likely to report higher rates of religious attendance and religious salience. Survey respondents in the telephone mode were also less likely to report being atheist than those who took the survey online.