I have illustrated this point from a detailed critique of a study conducted by Irene Jones. Murdock and Phelps (1973) surveyed adolescents (322 girls, 299 boys) about their lifestyles. They had designed separate questionnaires for boys and for girls, with some identical items and some which were different. When their original research instruments were scrutinised by Irene Jones (1974) it was clear that preconceptions, of a stereotyped sort, had shaped the data collection instruments. Murdock and Phelps gave Jones unrestricted access to all their unpublished data, so she could scrutinise all their instruments, analyses, and reasoning. One set of questions offered a variety of adolescent roles that the respondents could choose to identify with. Both sexes were offered ‘good pupil’, ‘rebel’, ‘ritualist’, ‘good bloke/good friend’, and ‘pop fan’. Boys were also offered ‘street peer’, ‘sports fan’, ‘boyfriend’, and ‘natural leader’. These were not offered to girls. Instead girls were offered ‘homemaker’, ‘tomboy’, ‘girl friend’ and ‘fashion follower’. Girls could not choose to say they were leaders, or hung out on the streets, or were sports fans. Boys could not choose to be home-centred (DIY or car mechanics with Dad, building model planes, gardening, cooking, ham radio…), or to be fashion followers, or to be ‘sissies’. The research instrument itself polarised the two sexes in stereotyped ways. The two questionnaires ensured that the results of the research revealed a gulf between the leisure patterns of teenage girls and boys, with boys out on the streets following sport and girls at home trying on each other’s clothes. By restricting the choices, the opportunity to find out how many female leaders and sports fans, or male home-bodies and fashion followers, was lost. It is likely that the vast majority of the adolescents would have claimed affinity with the stereotypes Murdock and Phelps expected, but because they built them in to their instruments, it will
never be possible to explore how many ‘home boys’ and ‘street girls’ there were.
Inside the detailed wording of the items there was further sexism. The male sports fan is good at sport, watches TV sport and goes to matches. His clothes and cleanliness are not mentioned. The tomboy likes swimming and gym but ‘does not like dressing up and would rather wear her old jeans all the time’. These are not equivalent. There is no equivalence between liking sport being closely associated with being unfashionable for girls, and a passion for sport being unmarked for fashion and sexual attractiveness for boys. It was equally stereotypical that the questionnaire nowhere provided for a boy to mention a passion for clothes or fashion. Irene Jones showed that the preconceptions about adolescent sex roles held by Murdock and Phelps had produced a stereotyped pair of questionnaires which were bound to produce a polarised set of results. Of course, it is possible that there were no young men who were fashion followers, or homeboys or sissies; and no young women who were in street gangs, or loved sport and fashion. However, because the full range of choices was not offered to both sexes, we will never know.
Exactly similar criticisms can be levelled at the research on social mobility. Hopper’s (1981) study of the ‘personal and interpersonal consequences of social mobility’ (ibid.: 13) is permeated with similar unexamined sexist (and heterosexist) assumptions. Hopper wanted to see how men compared themselves to their reference groups, and has a series of questions about other males in his informants’ lives. He took it for granted that no one was gay, and that the reference groups were all male. Thus, Hopper asked his informants, men in their 30s in 1965-6, about their ‘friends’ (explicitly men), who had ‘wives’ (ibid.: 125). In the retrospective questions about school, the men were asked about their ‘group of friends’ (males) and the ‘girls’ who they had ‘gone out with’ (ibid.: 255). When Hopper asked about relatives he focused on ‘their brothers and brothers-in-law’ (ibid.: 118).
The central point here was that if a researcher asks women about housework, or the quality of nursery provision, while questioning men about DIY and the quality of the railway network, it is not a legitimate finding that women hold views about nurseries and men about railways. Only if both sexes are asked about both topics, and answer differently, can the researcher legitimately report a sex difference. Many researchers in the 1950s, 1960s and 1970s would ask women if their husbands influenced how they voted, not ask men if their wives influenced them, and then report that women deferred to their husbands in political matters.
Most of the feminist criticism of the research methods current in sociology before 1980 focused on the research question, the sampling
and the instruments. There were fewer published critiques of the sexist nature of analytic processes and the ways in which writing up and publication were accomplished. This is partly because these stages were, in the 1945-80 period, usually regarded as non-problematic, and private. Very little self-conscious reflection on these topics was published by anyone, leaving less space for feminist critiques. The growth of confessional and reflexive accounts of all stages of the research process has made these topics more visible to feminist critiques.