Comparison of a self-administered questionnaire and a telephone interview of 146 Danish waste collectors.

Methods to collect epidemiologic data include, among other means, self-administered questionnaires and personal and telephone interviews. In some studies, these data collection methods are used simultaneously. However, little is known about the comparability of the data obtained, thus, the validity of pooling data. A self-administered questionnaire was completed by 146 waste collectors, who participated in a telephone interview 3 months later. The responses to health questions were compared. Agreement percentages and kappa values were calculated. An answer pattern shift was detected in a subgroup of questions with five answer options. Grouping the answer options diminished the shift in the answer pattern. The questions in a yes/no format generally showed no shift in the answer pattern. The study indicates that caution should be exercised when pooling data collected by self-administered questionnaire and telephone interview in epidemiological research. Some types of questions seem to be sensitive to the data collection method.