Within social sciences funding there is increasing emphasis on the reanalysis of existing data. Entire funding streams are dedicated to secondary analysis – though usually the focus is on quantitative datasets. Nonetheless, many researchers continue to seek opportunities to collect new data. While of course much can be learned from collecting new data, from new participants, I would also like to take a moment to reflect upon the virtues of revisiting and reanalysing our own existing data.
Most of us collect huge amounts of primary data throughout our careers yet we rarely revisit that data with the benefit of hindsight. The pressure to move on to the next funded project, means that we often have ‘data wastage’ whereby data are left unpublished and sometimes even un-analysed.
Lately, I have started to re-read interview transcripts that I collected over many years, from different migration projects, and I have found this a surprisingly rewarding experience. In particular, by combining different qualitative datasets, I have managed to generate new insights which have helped me to develop my understanding of complex issues. This process of re-visiting and re-analysing has resulted in two new publications.
In a recent paper on highly skilled women migrants, I brought together qualitative data collected as part of two separate studies – one on Irish migrants carried out in 2013 and the other on Polish migrants carried out in 2014 – and reanalysed them together as a combined corpus of data. Having previously looked at these young, highly qualified/ skilled Irish and Polish migrants separately, bringing them together in a comparative way allowed me to see new trends across the enlarged dataset. This comparative approach also helped me to overcome the ethnic exceptionalism which tends to predominate in migration research, much of which focuses on only one national or ethnic group.
Responding to on-going calls to look beyond the ethnic lens in migration research, I reanalysed my data through cross cutting themes, including for example, parenthood, age, professional status, transnational caring responsibilities, to reveal similarities across as well as differences within the data. Rather than specificities of Irishness or Polishness, factors such as having school-age children or ageing parents in the country of origin, proved particularly significant in shaping on-going decision making about staying, returning or moving on elsewhere. However, that is not to deny the ways in which specificities of migration status remain salient in migration research.
In another recent paper, Umut Erel and I have combined and reanalysed our separate corpus of qualitative migration data collected over nearly 20 years. These data, which included interviews with different groups of EU and non-EU citizens, enabled us to explore the impact of immigration regimes and citizenship status on mobility and settling opportunities.
In that paper, faced with a large corpus of rich qualitative data including over 100 in-depth interviews, we decided to adopt a case study approach. This method works well in qualitative research and allowed us to focus on particular themes such as citizenship through the rich and deep narratives of particular participants. In addition, combining data from Third country nationals and EU citizens was especially informative in the run up to Brexit since EU nationals may now be confronted with new challenges previously associated with non-EU migrants.
Revisiting data is not an excuse for ‘salami slicing’. This is certainly not an easy way of squeezing another publication out of a well exploited vein of materials. I do not pretend this is easy or quick. In both of the examples described here, a new analysis was undertaken of previously separate and distinct sets of qualitative data. Each re-analysis took considerable time as it meant engaging with or ‘revisiting’ an enlarged set of interviews, collected over many years.
Limitations should also be acknowledged. For example, participants may have been recruited using different sets of criteria, according to distinct project objectives and specific research questions. The temporality and spatiality of the data also need to be recognised. The geographical place and time period when interviews were conducted shapes, at least in part, the kind of data which have been collected. However, unlike archived qualitative data collected by other research teams, one advantage of revisiting and reusing one’s own data or working on combined data with colleagues, is that the research team also have some memory of when, how, and from whom the data were collected. These particularities of the data need to be fully acknowledged in any new outputs.
Nevertheless, while context is important, that does not preclude the emergence of new insights over the passage of time. Moveover, it is remarkable how often qualitative migration researchers ask similar kinds of questions to participants. Hence, data from different migration projects may be more comparable than expected.
Thus, in a climate where research funding is increasingly difficult to achieve, I would urge us to consider what we can learn from revisiting our own data. Working on combined sets of existing qualitative data either alone or in collaboration with colleagues, can provide abundant opportunities for interesting and insightful analysis generating new ways of understanding as well as providing opportunities to work beyond the limits of separate and particular datasets.
Louise Ryan is Professor of Sociology and Co-Director of the Migration Research Group at the University of Sheffield. Her two recent publications, discussed in this blog are:
Erel, U., & Ryan, L. (2018). Migrant Capitals: Proposing a Multi-Level Spatio-Temporal Analytical Framework. Sociology, https://doi.org/10.1177/0038038518785298
Ryan, L. (2018). Narratives of Settling in Contexts of Mobility: A Comparative Analysis of Irish and Polish Highly Qualified Women Migrants in London. International Migration, https://onlinelibrary.wiley.com/doi/abs/10.1111/imig.12493