Effects of Interview Mode on Self-Reported Well-Being

This study explores the differential effects of face-to-face interviewing and web-survey self-interviewing on subjective well-being questions. The analysis employs individual data from traditional surveys (e.g. the World Values Survey, the European Social Survey, and other representative national surveys) and data obtained from a web survey posted at WageIndicator websites in more than 30 countries. Web survey has become a very attractive tool as it allows access to respondents at a lower cost and large numbers relative to traditional surveys that are conducted via phone, mail, or on a face-to-face basis. The WageIndicator is a labor market oriented survey that also includes several subjective well-being indicators. In the first step we contrast the characteristics of samples resulting from different modes. Second, controlling for demographic characteristics, we test the hypothesis that online surveys decrease measurement error in sensitive questions created by the presence of an interviewer (e.g. the underreporting of socially undesirable behavior). We show that mostly in advanced economies respondents tend to report higher subjective well-being levels in the presence of an interviewer, while in developing and former communist countries the presence of an interviewer generates a downward bias in the reported well-being levels. The paper also contributes to the ongoing debate on web survey data quality, reliability, and validity for scientific use. It demonstrates how social sciences can benefit from the use of web survey data in order to overcome the limits of traditional information sources.