主講人：杜素豪、于若蓉、洪琮瑋
時間：2020.06.30 下午2時
地點：中研院人社中心第一會議室
ABSTRACT
In order to reduce cost and to reach more respondents after a few waves of data collection, mixed-mode surveys have widely adopted in panel studies. The respondents are, however, self-selected and unlikely to be randomly assigned to different survey modes. To distinguish the mode effects from the selection effects, therefore, has become a challenge to survey methodologists. Among newly recommended statistical methods which can control the self-selection bias, we use propensity score analysis to examine the mixed-mode effects of a large-scale panel survey conducted in Taiwan. The data used in this study are from the Panel Study of Family Dynamics. From its first data collection to the 2016 survey, almost all questionnaires were completed by face-to-face interview. Starting with the 2018 survey, an internet self-administered questionnaire has been added. Using the data collected in 2018, we try several matching approaches of propensity score analysis, such as one-to-many matching, optimal matching, and full matching, appropriate to examine the mode effects on items sensitive to social-desirability bias, including happiness, self-reported health, and various measures of life satisfaction. The findings are beneficial to our understanding of the mixed-mode effects associated with panel surveys.