Recent work has demonstrated that Canadians overreport church attendance at rates similar to respondents in the United States. Overreporting in the United States has been attributed to the importance of religious identity; causes of Canadian overreporting have not been examined. This
article draws upon Stryker's identity theory to explain why Canadian survey respondents overreport church attendance. The 2005 Canadian General Social Survey contains observed measures of attendance from both a conventional survey question and a time diary, allowing a direct and rigorous test
of the identity explanation. Findings suggest that rates of Canadian overreporting, at about 50 percent, rival rates of overreporting in the United States. Moreover, like overreporting in the United States, Canadian overreporting may be affected by an identity process during the survey interview.
Finally, implications for measuring religiosity in both countries are discussed.