Подпись: 63The choice of research question sets the agenda for any research proj­ect. In the past, many sociologists set themselves research questions which, when scrutinised by feminists, turned out to have been embedded in sexist assumptions and/or actually focused on only the male sex while purporting to be universal. If we take the research question from a well-known, respected, and much cited project con­ducted in the 1970s, this general point becomes clear. The authors pose their research question as follows: ‘Does the labour market objectively allow to the worker a significant measure of choice over his economic life? Does he subjectively perceive this as choice?’ Given the pattern of using male pronouns to cover both sexes, this could be a research question about workers of both sexes. However, it was not. The research question comes from Blackburn and Mann (1979: 2) characterising their study of 1,000 workers (actually 1,000 men), in Peterborough. The project is a study of semi-skilled men in a local labour market with a choice of jobs. It is an interesting read. However, it is not about workers. It is about male workers. Feminist sociologists needed to point out that failing to specify gender in the research question reduces clarity; specifying gender sharpens the research question.

The same point can be made about research on education and social mobility. Sandra Acker (1981), for example, highlighted how much of the empirical research in the sociology of education in Britain had been done on all-male samples. For example Hope asks ‘What did it feel like to be a child in the Scottish system in 1947?’ (1984: 19), Hopper (1981: 13) asked what are ‘the personal and interpersonal consequences of social mobility?’. Both men actually only studied males. Hope researched what is was like to be a boy in the Scottish system, and Hopper what the consequences of social mobility in England were for men. There is a relationship between the research question and the choice of a research setting or the way the sampling is done. Some of the most criticised sampling was done on social mobility issues.