We recently had the pleasure of presenting our work at the 7th ESRC Research Methods Festival, hosted by the University of Bath. And what an event it was! Over three days we joined social scientists from a huge range of disciplines and sectors and at different points in their research careers.
It was great to find that many of the methodological issues being discussed ‘spoke’ to our own research. Big Data, longitudinal methods and the qualitative / quantitative divide were all hotly debated. Jane Elliott’s keynote addressed the term ‘Big Data’, the digital revolution and the computational techniques being developed to analyse large corpuses of unstructured data. Jane highlighted the possibilities machine learning can bring to Big Data analysis, but also the important role of the researcher in determining clear research questions, and validating outputs. She concluded that rather than dividing qualitative / quantitative research further, computer-assisted methods can empower, and bring both groups together.
Time was also a focus: from Bartlett’s look at the diary methods as a means of capturing lives as they are lived; Gershuny’s and Sullivan’s consideration of time use diaries for understanding the rhythms and patterns of everyday activities; and Goodman’s work on tracking, over time, the experiences of those living and dying with dementia in care homes. Technologies are advancing many of these methods, and researchers are being afforded greater access to longitudinal datasets. CLOSER Discovery was one resource promoted at the Festival, an online resource enabling researchers to view and appraise data from eight of the UK’s leading longitudinal studies.
So where does our research fit with these issues? We presented with Anna Tarrant, Leverhulme Fellow at the University of Leeds. Anna’s work has featured on our site before (check it out here), and together we outlined our ongoing strategies for working with combined qualitative longitudinal datasets.
As we detailed in our presentation, we are likening our work to the stages of an archaeological excavation and are finding this helpful in thinking about how we access the data at different levels and in different ways. This is a rather apt metaphor for interrogating the large and complex ‘field’ of QLR since it invokes the notion of understanding the layers of the past, excavating the past through waves of data, and across the lifecourse.
This approach has taken several stages, with each ‘digging’ further into the dataset. We began with explorations akin to a ‘surface survey’ to evaluate the breadth and nature of the archive, and to organise it into a composite dataset. We then used ‘geophysical surveys’ to explore the landscape of the data without penetrating the surface. Here we employed a range of computer-assisted technologies to examine word frequencies; words or phrases that often used with other words or phrases; and word clusters. Using the outcomes of our different ‘surveys’, we identified ‘cases’ for further testing. In these shovel test pits we were able to use example cases to simultaneously compare computer-assisted corpus-oriented analyses with more conventional in-depth explorations.
Our next step is to move onto ‘deep excavations’. These will use the outcomes of our ‘surveys’ and ‘test pits’ to inform where and how we delve deeper into the detail of selected cases, with a focus on change and continuity across layers of time. It is here, as Jane suggests, that we will start to ask specific and focused questions of our data.
We have found that while quantitative methods for analysing large corpuses can provide insight into the texture of a qualitative dataset, concurrent ‘thick’ readings can be used productively to expand meaning and understandings. ‘Big’ qualitative data analysis necessitates a combination of interpretive techniques and, with this, has the potential to bring qualitative and quantitative approaches into conversation. We hope to continue this conversation as we progress our research.
We hope those who attended our session enjoyed it: we are looking forward to sharing our work further. One such event will be an NCRM short course, which we hope to publicise soon. In the meantime, the National Centre for Research has a wide range of training and events to choose from across the UK.