Q.An investigator wants to study people's attitude towards welfare programmes by circulating a questionnaire via the Internet. Is this study likely to reflect the views of the general population accurately? Why or why not?

Sampling bias occurs whenever the group of people a researcher actually manages to survey systematically differs from the wider population the researcher intended to describe, so that the survey's conclusions end up describing the respondents rather than the population as a whole. This is a genuine risk with several of the survey techniques discussed in the chapter. An internet-circulated questionnaire, for instance, can only ever reach people who have internet access, who are comfortable using it, and who happen to choose to respond, three separate filters, each of which can quietly exclude large sections of the wider population, such as people without reliable internet access, older people less comfortable online, or simply the many people who see the questionnaire but never bother to fill it in. The resulting sample of actual respondents may therefore look nothing like the general population in terms of age, income, education, location or even the very attitudes the survey is trying to measure, undermining the survey's cl …