【摘要】This study describes the existing problems and mistakes that may cause misunderstanding or even misleading in current analyses of questionnaire survey data and suggests methods to do exploratory data analyses and select related variables for further inference.