How many research methods are enough?

NCRM news
Graham Crow, University of Edinburgh
Geometric shapes fitting into a puzzle as a metaphor for research methods that can stand alone or complement each otherGeometric shapes fitting into a puzzle as a metaphor for research methods that can stand alone or complement each other

For researchers who decide to use more than one method in a project, the question immediately arises of whether to treat two as sufficient or whether to go further. Just as single method studies can be perfectly adequate for many purposes, so two methods may be all that is required for others, while some venture into using three or more. I once asked Alan Bryman after a presentation he had made on his research into mixed methods what studies he would point to as exemplars of the field. After a brief period of reflection – long enough to indicate that he’d needed to think about it because he didn’t have a ready-made answer – he nominated Janet Finch and Jennifer Mason’s Negotiating Family Responsibilities (London: Routledge, 1993). Its combination of a survey of 978 survey respondents with qualitative interviews of 88 participants (who as part of the process engaged with vignettes) is indeed a model piece of research. In the ensuing conversation my own suggestion of Ray Pahl’s Divisions of Labour (Oxford: Basil Blackwell, 1984) was also considered, although the fact that it drew on data generated using a wide range of methods may have worked against its identification as a model for others to follow. Having the funding to support a large team of researchers who were in the field for the best part of a decade, Pahl’s study of the Isle of Sheppey drew initially upon ethnographic observation, informal interviews, more structured interviews, documentary research (involving newspapers both contemporary and archived), photography, and imagined futures essay-writing undertaken by young people close to the point at which they left school and entered the labour market. 

On the strength of the material that these methods allowed to be collected, Pahl was able to put some tentative but nevertheless controversial arguments about the changing nature of work into the public domain. However, the adoption of further methods of investigation transformed the account of Sheppey and its people that emerged. Oral history interviews about working for the Admiralty Dockyard in Sheerness prior to its closure in 1960, together with archival research going back to the nineteenth century, allowed a more historically-informed picture to emerge that contextualised the local culture. In addition, a survey of and expert interviews with local employers provided the points of view of firms which operated on the island (viewpoints which did not always square with those of employees which Pahl had collected previously). But it was the survey of 1 in 9 Sheppey households commissioned by Pahl that would transform the project, forcing the team to re-think not only the pervasive myth relating to the islanders’ long-standing residence (necessitated by the finding that many were recent arrivals) but also the myth, again pervasive, that people who became unemployed could ‘get by’ satisfactorily through involvement in the informal or hidden economy. The household survey provided compelling evidence that unemployment of an individual tended to have an adverse effect on the economic position of other household members as well, including effects across generations. The survey provided the foundation for the study’s main conclusion, that de-industrialization was producing social polarization, driving a wedge between prosperous households that were ‘work-rich’ and households that were ‘work-poor’ and downwardly mobile.

If the four different types of interviewing (informal, structured, elite and oral history) are counted separately, the study employed no fewer than nine research methods in total. Each of these methods made a distinct contribution to rounding out the account of the place and its people that Pahl and his team produced, and for each method the case can be made that the project would have been the poorer without it. Of course, the constraints of time, budget and personnel within which we operate normally are such that our research designs will be less ambitious than Pahl’s. Who can envisage spending several years in the field, aided by a team of researchers, and with the resources to approach an issue from diverse methodological perspectives? Yet it remains the case that there may be some value in considering whether using an additional method in our projects might allow access to another dimension that would otherwise be missed. In many cases, quite possibly most, potential benefits of this kind will be outweighed by the costs of incorporating another method, including the very real risk that different methods will throw up discrepant findings that are difficult to incorporate. But the Sheppey study’s results reveal that such challenges can act a spur to the pursuit of a deeper understanding of the subject under investigation. Pahl’s use of social network analysis in a subsequent project indicates that his time on Sheppey in no way dimmed his appetite for methodological experimentation, including adopting new methods.


Graham Crow is Professor of Sociology and Methodology at the University of Edinburgh, where he has worked since 2013. His publications in the field of research methods include his 2021 book Community Studies in the Bloomsbury Research Methods series, which discusses Pahl’s Sheppey study as an exemplar of community studies research. The slides for Graham’s talk on this topic given at the Research Training Centre at the University of Edinburgh are available here.