Abstract
The effects of question form and interviewer style on the accuracy of reporting automobile accidents in a household interview are studied. A professional style interview is superior to an impersonal style interview. some empirical evidence on the nature of the interviewer-respondent interaction process are added.