ANNOYANCE ASSESSMENTS IN POSTAL SURVEYS USING THE 5-POINT AND 11-POINT ICBEN SCALES: EFFECTS OF SCALE AND QUESTION ARRANGEMENT

摘要

Background: The type of noise annoyance scale and its presentation (question location, order of answer alternatives etc.) in questionnaires or interviews as well as other contextual factors potentially affect annoyance responses and explained variance in socio-acoustic surveys. Methods: By means of a balanced experimental design, we investigated the effect of type of scale (5-point vs. 11-point ICBEN scale), presentation order of answer alternatives (ascending vs. descending) and question location (early vs. late) within a postal questionnaire on annoyance responses and model fit characteristics. The survey was carried out on a stratified sample of 1217 residents affected by road traffic noise between 50 und 70 dB(A) Ldn. Results: Early appearance of annoyance questions was positively associated with a higher annoyance score as well as with a higher probability of reporting "high annoyance". The order of answer alternatives had no effect. Annoyance responses on the 11-point scale were slightly higher and yielded larger R-squareds in linear exposure-effect models than responses on the 5-point scale, as determined by bootstrap resampling. Conclusion: The placement of annoyance questions within a postal questionnaire has an effect on the level of reported annoyance, with early appearing annoyance questions resulting in higher annoyance scores. The order of answer alternatives has no effect. Statistical exposure-annoyance models fit slightly better using the answers from the 11-point scale. Whether these findings also apply to face-to-face or telephone interviews remains to be investigated.