Research on the social and psychological effects of mobile phone communication primarily is conducted using self-report survey measures. However, recent studies have suggested such measures of mobile phone communication use contain a significant amount of measurement error. This study compares the frequency of mobile phone use measured by self-report questions with error-free log data automatically collected through an Android smartphone application. We investigate the extent to which measurement non-random error exists in the self-report questions and the predictors of this error. The data were collected from a sample of 310 Android phone users residing in Japan. Our analysis shows that users generally overreport their frequency of mobile communication and that over-estimation is better predicted by proxy measures of social activity than demographic variables. We further show an example of how over-reporting can result in an overestimation of the effects of mediated communication on civic engagement. Finally, the value of behavioral log data in mediated communication research is discussed.
Mendeley saves you time finding and organizing research
Choose a citation style from the tabs below