Social Desirability Bias and Mode Effect in the Case of Voting Behavior

Response distribution and data quality are influenced by different mode of data collection. Such consequences are seen as mode effect. One significant influence of mode effect on data quality is social desirability bias, which is related to whether an interviewer is involved, pace of cognition process during interviews, sensitivity of survey questions, etc. For example, social desirability bias is more likely to occur in face-to-face interview when compared to telephone interview. Slower pace and the development of rapport in the former encourage respondents to think thoroughly and provide a socially desirable answer. However, previous studies on social desirability of mode effect did not obtain consistent findings. The problem of social desirability bias is found to be more severe in telephone interview than in face-to-face interview when dealing with voting turnout (Díaz de Rada, 2011; Voogt and Saris, 2005). The issue of such responding bias with the consideration of mode effect is worth further examination.

This study aims to examine mode effect on social desirability bias by comparing CATI and face-to-face interviews on the voting turnout of the 2004 president election in Taiwan. Data are drawn from two national representative sample conducted in the same year after the election to eliminate recall error, with face-to-face survey data from Taiwan Social Change Survey and CATI data from Taiwan Social Image Survey. After weighting by population characteristics, socio-demographic variables will be compared first to examine the dis/similarity of the two samples. Voting behavior, political attitudes and other related variables will be included in the multivariate analysis. Conclusion and discussion will be provided.