Three major methods of survey research, face-to-face interviews, telephone interviews, and mail questionnaires, are compared with respect to the quality of the data. The literature on experimental comparisons of these methods is reviewed, and the effects of the mode of data collection on aspects of data quality are examined. The effects of the data-collection method on research results are also examined with a focus on the consequences for the relations among variables and emerging empirical models. The meta analysis is followed by a field experiment with 762 responses. Meta analysis detected small differences between the modes, suggesting a dichotomy between modes with and without an interviewer. The field experiment found the lowest response rates for the face-to-face survey, with more item nonresponse in the mail survey but more self-disclosure through the mail. The mail survey was slightly superior in reliability and scalability. Results suggest that interviewer training should be adapted to the changes in data-collection mode. Five figures and 33 tables present meta analysis and survey findings. Three appendixes contain a bibliography, the questionnaire content, and marginal distributions of background variables.