Vibration in dwellings from road and rail traffic — Part II: exposure–effect relationships based on ordinal logit and logistic regression models

The 1998 Norwegian Socio-vibrational Survey with 1503 respondents from 14 study areas was undertaken in the autumns of 1997 and 1998. Telephone interviews were used for obtaining peopleʹs reactions to vibrations in dwellings while in-dwelling vibration values were calculated for 1427 of these respondents. The response rate was 50%. The calculations of the vibration values were based on measurements and calculations by using a semi-empirical vibration prediction model. The logarithmic transformation of the calculated statistical maximum vibration value vw,95 as defined in the new Norwegian Standard NS 8176, was selected as exposure measure. There were no significant differences in reactions to vibrations from different sources. An estimated exposure–effect ordinal logit model for degrees of annoyance as a function of the vibration exposure measure shows good fit and narrow statistical error bands. Further studies are needed to determine whether the exposure measure should also include the number and duration of vibration events.