Does low willingness to respond introduce a bias? Results from a socio-epidemiological study among young men and women

The aim of this study was to investigate if willingness to respond can intr oduce bias in a population survey and to study the socio-economic character istics of those with high and low willingness to respond. Two groups of fin al-year pupils at compulsory school were surveyed five years after leaving school, at the age of 21. The data were collected during class meetings, by mailed questionnaires, telephone and personal interviews. The response rat es in the studies were proportionally high - about 98% and 90% respectively . In both surveys there were more men compared to women who were late respo ndents or non-respondents. Long-term unemployment and lack of upper seconda ry school for both men and women, as well as living with a partner for men, corresponded to low willingness to respond to a survey. A higher alcohol c onsumption - between 1.5 to 6.5 times - was observed among those with lower willingness to respond. Even a low rate of non-response can introduce a se rious bias and produce uncertain results in, for example, studies on the co nsequences of youth unemployment on alcohol consumption.