Using computer-assisted survey instruments instead of paper and pencil increased completeness of self-administered sexual behavior questionnaires

Abstract Objectives To compare the data quality, logistics, and cost of a self-administered sexual behavior questionnaire administered either using a computer-assisted survey instrument (CASI) or by paper and pencil in a primary care clinic. Study Design and Setting A self-administered sexual behavior questionnaire was administered to 16–29 year olds attending general practice. Questionnaires were administered by either paper and pencil (paper) or CASI. A personal digital assistant was used to self-administer the CASI. Results A total of 4,491 people completed the questionnaire, with 46.9% responses via CASI and 53.2% by paper. Completion of questions was greater for CASI than for paper for sexual behavior questions: number of sexual partners [odds ratio (OR), 6.85; 95% confidence interval (CI): 3.32, 14.11] and ever having had sex with a person of the same gender (OR, 2.89; 95% CI: 1.52, 5.49). The median number of questions answered was higher for CASI than for paper (17.6 vs. 17.2; P