Guest blog # 8: Dr Sarah Wilson: Using qualitative secondary analysis as a tool of critical reflexivity

Our guest post today is by Sarah Wilson, a Senior Lecturer in Sociology in the School of Applied Social Science at the University of Stirling. Sarah’s research interests are in the sociology of families, relationships and personal life, with a methodological focus on developing visual, audial and artistic qualitative research. In this post, Sarah reflects on her qualitative secondary analysis of data from the Timescapes ‘Siblings and Friends’ project, a longitudinal dataset with which we are also working, and how this process prompted reflection on her own research practices.

This post draws on Sarah’s 2014 article in Sociological Research Online, ‘Using secondary analysis to maintain a critically reflexive approach to qualitative research’ which you can read here: http://www.socresonline.org.uk/19/3/21.html

 

 Using qualitative secondary analysis as a tool of critical reflexivity

Maintaining a critical, reflexive approach to research when engaging in specialised work is not easy. Partly because of the need to convince funders of their expertise, researchers often focus on relatively circumscribed areas of inquiry, with samples drawn from particular social groups.

My own research has focused on samples characterised as ‘vulnerable’; notably young people affected by parental substance misuse or living ‘in care’. Often this work has been located within more ‘applied’ approaches to social research, and influenced by funders’ concerns. Such work is valuable. However, the segregation often maintained between research with young people from more ‘vulnerable’ and more ‘ordinary’ backgrounds may reinforce perceptions that the experiences, values and aspirations of members of each ‘category’ are distinct. As Law (2009) argues, research is ‘performative’, helping to re-produce and reinforce perceptions of social groups. In the current political context, such distinctions may even implicitly reinforce the stigmatisation of ‘troubled families’. As such, there is a need to find ways to subject one’s own research practice to scrutiny.

To better situate my previous research, I engaged in qualitative secondary analysis of the longitudinal Timescapes ‘Siblings and Friends’ (SAF) study to prepare for a new project with ‘looked after’ young people: Young people creating belonging: spaces, sounds and sights (ESRC RES-061-25-0501). The idea was to reflect on my own approaches, and previous framings of interview questions in the light of the very rich SAF project data which involved predominantly ‘ordinary’ young people from across the UK. This proved to be an illuminating, if demanding, process that prompted further thought about both projects.

Importantly, this analysis highlighted significant commonalities between the experience of those included in ‘ordinary’ and ‘vulnerable’ samples. Notably, the SAF data included several accounts of strained family relationships, of parental mental ill-health and of undesirable housing conditions that suggested family circumstances comparable to those in my previous work on parental substance misuse. However, the SAF interview questions situated violence outside of the home. As Gillies (2000) argues, even where ‘difficult’ accounts within ‘ordinary’ samples are identified, they are often not written up. As such, the complexity and pain within ‘ordinary’ families may be under-estimated in research, and potentially more easily obscured within political discourse. Similarly, the everyday ambiguity and minor conflicts associated with ‘ordinary’ siblings and parents sharing limited space may be downplayed. Such ambiguities and tensions led several SAF respondents to seek out friends’ homes, or private corners of their own, to escape from family life at least for a time. I had previously associated such strategies with young people affected by parental substance use, many of whom often spent time at friends’ houses. However, this analysis suggested a more nuanced understanding of the importance to the latter group of employing strategies that could be presented as ‘ordinary’ teenage practices. The process of secondary analysis also highlighted uncomfortable omissions from my previous research in which, for various reasons, greater emphasis was placed on the respondents’ own potential substance use, than on their school work and employment aspirations. The predominance of such concerns in the SAF accounts led me to worry that my own research had reflected and performed perceptions of education as less important to ‘vulnerable’ than to ‘ordinary’ young people.

In conclusion, qualitative secondary analysis is a ‘labour-intensive, time-consuming process’ that Gillies and Edwards (2005: para24) compare to primary data collection. However, it presents a useful tool to subject assumptions built up over a specialised research career to scrutiny.

 

References 

Gillies, V. (2000) ‘Young people and family life: analysing and comparing disciplinary discourses’, Journal of Youth Studies, 3(2): 211-228

Gillies, V. and Edwards, R. (2005) ‘Secondary analysis in exploring family and social change: addressing the issue of context’, Forum: Qualitative Social Research, 6(1): art 44.

Law, J. (2009), ‘Assembling the World by Survey: Performativity and Politics’, Cultural Sociology, 3, 2, 239-256.

Wilson, S. (2014) ‘Using secondary analysis to maintain a critically reflexive approach to qualitative research’ Sociological Research Online, 19(3), 21 http://www.socresonline.org.uk/19/3/21.html

 

 

 

Guest post #7, Dr Gregor Wiedemann: Computer-assisted text analysis beyond words

robot-507811_1920Dr Gregor Wiedemann works in the Natural Language Processing Group at Leipzig University. He studied Political Science and Computer Science in Leipzig and Miami. In his research, he develops methods and workflows of text mining for applications in social sciences. In September 2016, he published the book “Text Mining for Qualitative Data Analysis in the Social Sciences: A Study on Democratic Discourse in Germany” (Springer VS, ISBN 978-3-658-15309-0).

In this blog, he discusses computational textual analysis and the opportunities it presents for qualitative research and researchers. 

Computer-assisted text analysis beyond words

In our digital era, amounts of textual data are growing rapidly. Unlike traditional data acquisition in qualitative analysis, such as conducting interviews, texts from (online) news articles, user commentaries or social network posts are usually not generated directly for the purpose of research. This huge pool of new data provides interesting material for analysis, but it also poses qualitative research with the challenge to open up to new methods. Some of these were introduced in blog post #4. Here, computer-assisted text analysis using Wordsmith and Wordstat were discussed as a means of allowing an ‘aerial view’ on the data, e.g. by comparative keyword analysis.

Despite the long history of computer-assisted text analysis, it has stayed a parallel development with only little interaction with qualitative analysis . Methods of lexicometric analysis such as extraction of key words, collocations or frequency analysis usually operate on the level of single words. Unfortunately, as Benjamin Schmidt phrased it, “words are frustrating entities to study. Although higher order entities like concepts are all ultimately constituted through words, no word or group can easily stand in for any of them” (2012). Since qualitative studies are interested in the production of meaning, of what is said and how, there certainly are overlaps with lexicometric measures, but nonetheless their research subjects appear somewhat incompatible. Observation of words alone without respect to their local context appears as rough simplification compared to a hermeneutic close reading and interpretation of a text passage.

The field of natural language processing (NLP) from the discipline of computer science provides a huge variety of (semi-) automatic approaches for large scale text analysis, and has only slowly been discovered by social scientists and other qualitative researchers. Many of these text mining methods operate on semantics beyond the level of isolated words, and are therefore much more compatible with established methods of qualitative text analysis. Topic models, for instance, allow for automatic extraction of word and document clusters in large document collections (Blei 2012). Since topics represent measures of latent semantic meaning, they can be interpreted qualitatively and utilised for quantitative thematic analysis of document collections at the same time. Text classification as a method of supervised machine learning provides techniques even closer to established manual analysis approaches. It allows for automatic coding of documents, or parts of documents such as paragraphs, sentences or phrases on the basis of manually labelled training sets. The classifier learns features from hand coded text, where coding is realised analogously to conventional content analysis. The classifier model can be seen as a ‘naïve coder’ who has learned characteristics of language expressions representative for a specific interpretation of meaning of a text passage. This ‘naïve coder’ then is able to process and code thousands of new texts, which explicitly opens the qualitative analysis of categories up to quantification.

In my dissertation study on the discourse of democratic demarcation in Germany (Wiedemann 2016), I utilised methods of text mining in an integrated, systematic analysis on more than 600,000 newspaper documents covering a time period of more than six decades. Among others, I tracked categories of left-wing and right-wing demarcation in the public discourse over time. Categories were operationalised as sentences expressing demarcation against, or a demand for, exclusion of left-/right-wing political actors or ideologies from the legitimate political spectrum (e.g. “The fascist National Democratic Party needs to be banned” or “The communist protests in Berlin pose a serious threat to our democracy”). Using automatic text classification, I was able to measure the distribution of such qualitatively defined categories in different newspapers between 1950 and 2011. As an example, the following figure shows relative frequencies of documents containing demarcation statements in the German newspaper, the Frankfurter Allgemeine Zeitung (FAZ).
gregor-1Distribution indicates that demarcation towards left-wing actors and ideology long-time superseded right-wing demarcation. Soon after 1990, the latter became the primary discourse subject of threats of German democracy. The enormous benefit of automatic classification is that it allows for easy comparison of publications (e.g. other newspapers) or relations with any other category. For instance, the distribution of “reassurance of democratic identity”, a third category I measured, strongly correlates with right-wing demarcation, but not with left-wing demarcation. Such a finding can be realised only by a combination of the qualitative and the quantitative paradigm.

While computer-assisted methods support qualitative researchers clearly in their task of retrieving “what” is being said in large data sets, they certainly have limitations on the more interpretive task of reconstructing “how” something is said, i.e. the characterisation of how meaning is produced. It is an exciting future task of qualitative research to determine how nowadays state-of-the-art NLP methods may contribute to this requirement. In this respect, computational analysis extends the toolbox for qualitative researchers by complementing their well-established methods. They offer conventional approaches new chances for reproducible research designs and opportunities to open up to “big data” (Wiedemann 2013). Currently, actors in the emerging field of “data science” are a major driving force in computational textual analysis for social science related questions. Since I repeatedly observe lack of basic methodological and theoretical knowledge with respect to qualitative research in this field, I look forward to a closer interdisciplinary integration of them both.

Further reading

Blei, David M. 2012. “Probabilistic topic models: Surveying a suite of algorithms that offer a solution to managing large document archives.” Communications of the ACM 55 (4): 77–84.

Schmidt, Benjamin M. 2012. “Words alone: dismantling topic models in the humanities.” Journal of Digital Humanities 2 (1). Url http://journalofdigitalhumanities.org/2-1/words-alone-by-benjamin-m-schmidt.

Wiedemann, Gregor. 2013. “Opening up to Big Data. Computer-Assisted Analysis of Textual Data in Social Sciences.” Historical Social Research 38 (4): 332-357.

Wiedemann, Gregor. 2016. Text Mining for Qualitative Data Analysis in the Social Sciences: A Study on Democratic Discourse in Germany. Wiesbaden: Springer VS, Url: http://link.springer.com/book/10.1007%2F978-3-658-07224-7.

Guest post #6, Nick Emmel: Revisiting yesterday’s data today

past-present-futureToday we welcome Dr Nick Emmel as our guest blogger. Nick has been investigating social exclusion and vulnerability in low-income communities in a city in northern England since 1999. The research discussed in this blog, Intergenerational Exchange, was an investigation of the care grandparents provide for their children. This was a part of Timescapes, the ESRC’s qualitative longitudinal research initiative. More details of this research are available at http://www.timescapes.leeds.ac.uk/research/intergenerational-exchange.html.

In this thought provoking post, Nick reflects on his experiences of revisiting qualitative data, and the ways in which new interpretations and explanations are generated over time. 

 

Revisiting yesterday’s data today 

I have recently finished writing a paper about vulnerability. This is the third in an ongoing series of published papers; the first published in 2010 and the second in 2014 (Emmel and Hughes, 2010; 2014; Emmel, 2017). Each elaborates and extends a model of vulnerability. All three are based on the same data collected in a qualitative longitudinal research project, Intergenerational Exchange, a part of Timescapes and its archive. The second and third paper draw on newly collected data from subsequent research projects as well. In this blog I want to explore how interpretation and explanation are reconstituted and reconceived through engagement with these new data and theory, considering some methodological lessons in the context of qualitative longitudinal research.

At first sight the narratives told us about poverty, social exclusion, and the experiences of grand parenting by Bob and Diane, Ruth, Sheila, Geoff and Margaret, and Lynn, which populate these three papers seem fixed, even immutable. After all, I am still using the same printed transcripts from interviews conducted between 2007 and 2011, marked up with a marginalia of memos and codes in my micrographia handwriting, text emphasised with single and double underlines in black ink. But each time I get these transcripts out of the locked filing cabinet in my office I learn something new.

To start with there are the misremembered memories of what is actually in the transcripts. Many of the stories our participants tell, Geoff and Margaret’s account of the midnight drop, Sheila bathing her kids in the washing machine, or Lynn walking into the family court for the first time, I have retold over and over again. In their retelling details have been elaborated, twisted, and reworked to make better stories so my students, service deliverers, and policy makers will think a little harder, I hope, about powerlessness, constrained powerfulness, and ways in which excluded people depend on undependable service delivery. In this way they are no different to the original stories, neither truth nor untruth, but narrated for a purpose, to describe experience in qualitative research. Getting the detail and emphasis right is important. The participants know their lived experience far better than I do. Re-reading the transcripts, these stories are reattached to their empirical moorings once again. But this is only the start of their reanalysis.

Rereading may confirm empirical description but past interpretations are unsettled by new empirical accounts. New knowledge has the effect, as Barbara Adam (1990:143) observes, of making the ‘past as revocable and hypothetical as the future’.  In the most recent of the three papers the apparently foundational role of poverty elaborated in our first paper is reinterpreted. New data from relatively affluent grandparents describe the barriers they face in accessing services and the ways in which these experiences make them vulnerable. This knowledge has the effect of reconstituting the original transcripts, shifting attention away from the determining role of poverty to relationships with service providers in which poverty may play a generative part. These data evoke new interpretations. But it is not only new empirical accounts that reshape this longitudinal engagement, new ideas are at play.

In this blog I have suggested that new empirical accounts change how we understand and interpret existing data. To ascribe reinterpretation only to these insights is not enough however. Explanations rely on more than reconstructing empirical accounts in the light of new insight. For a realist like me theories guide the reading of the original transcripts and the collection of new data. Theories are practical things, bundles of hypotheses to be judged and refined empirically. We started with a theory about time as a chronological progression of events, as is explained in the first paper. For our participants, they noticed little difference as recession merged with recession all the way back to the closure of the estate’s main employer in 1984. This theory was found wanting when we came to looking at young grandparenthood and engagement with service provision in the second paper. A refined theoretical account of the social conscience of generational and institutional time supported explanation. These theories, like the empirical accounts of the social world they are brought into a relation with, are revocable and only ever relatively enduring.

To paraphrase the Greek philosopher Heraclitus, no researcher ever steps into the same river twice, for it is not the same river and it is not the same researcher. Revisiting yesterday’s data today reminds us of these methodological lessons in qualitative longitudinal research.

References

Adam, B (1990) Time and social theory Polity Press, Cambridge.

Emmel, N. (2017) Empowerment in the relational longitudinal space of vulnerability. Social Policy and Society. July.

Emmel, N. & Hughes, K. (2010) “‘Recession, it’s all the same to us son’: the longitudinal experience (1999-2010) of deprivation”, 21st Century Society, vol. 5, no. 2, pp. 171-182.

Emmel, N. & Hughes, K. (2014) “Vulnerability, inter-generational exchange, and the conscience of generations,” in Understanding Families over Time: Research and Policy, Holland J & Edwards R, eds., Palgrave, Basingstoke.

Image source: Fosco Lucarelli ( https://www.flickr.com/photos/fosco/3915752142/).

 

Guest post #5, Sue Bellass: The challenges of multiple perspectival QL analysis

 

IMG_7916-Edit-800x800Our guest post today is by Sue Bellass, a PhD student in the School of Nursing, Midwifery, Social Work and Social Sciences at the University of Salford. Her thesis, which she is due to submit in August, has been exploring how intergenerational families are affected by young onset dementia over time.

In this post, Sue shares in detail her approach to analysing data over time, from multiple perspectives. The process has been complex and challenging, but has also brought creativity and freedom – and ultimately a deeper understanding of the lived experience of young onset dementia.

If you would like to know more about Sue’s research, contact her by email: s.bellass@edu.salford.ac.uk.

The challenges of multiple perspectival qualitative longitudinal (QL) analysis: a strategy created for an intergenerational study of young onset dementia

Although dementia is often perceived to be a condition that occurs in later life, around 1 in 20 people with dementia are below the age of 65 (Alzheimer’s Society, 2015). Over the last two decades there has been increasing interest in developing qualitative understandings of the experience of the condition in younger people; however, almost without exception existing studies have used cross-sectional designs, providing only a snapshot of life with an unpredictable, dynamic condition. For my PhD I decided to use a QL methodology to explore relationality over a twelve-month period by following five intergenerational families where one person had received a diagnosis of young onset dementia.

Since people with dementia are a marginalised, negatively positioned group (Sabat et al., 2011), I felt it was appropriate to democratise the research process to enable my participants to choose their preferred means of engaging with the study. This choice included the method of data collection (ethical approval was gained for interviews, audio/ video diaries, blogs and tweets) and, if participants opted for interviews, which family members would participate and where the interviews would take place.  Ultimately, 18 participants chose to be interviewed, 16 of whom were interviewed in pairs or larger family groups, with two preferring individual interviews. Interviews were conducted in three waves at months 0, 6 and 12.

Analysing the data set has been a challenging process. As Henderson et al. (2012) note, despite increasing interest in QL methods, methods of analysing and representing complex QL data sets have rarely been explicated. I experienced this as a mixed blessing; on the one hand, there is space for creativity, flexibility and freedom, on the other, there is room for doubt to flourish!  I have attempted to slice the data in different ways in order to interrogate the data set to best effect.  Inspired by Thomson (2010, 2014), I treated each family as a unique case and also aimed to create a cross-case analysis across the four generations represented in the families.

Example QL matrices

matrix_example

Initially I attempted to analyse the group interviews at the ‘family’ level, however it quickly became apparent that divergent accounts were being obscured.  Subsequently I took a multiple perspectival approach (Ribbens McCarthy et al., 2003), teasing apart individual experiences within the families, viewing them as cases within a case. For each person, I induced categories of experience then, to permit holistic re-engagement, organised the raw data in a time-ordered matrix across the three waves.

Then, again for each person, I created a longitudinal matrix adapted from Saldana (2003) to look for transitions and continuities, using motif coding, a form of coding which draws attention to recurring elements in experiences, and describing through-lines, a crystallisation of a participant’s change over time. Although it could be argued that such an approach may disguise intersubjective creation of meaning, I consciously retained a focus on relationality, creating spaces within the matrix to capture data on meaning-making processes over time. Finally I created an intergenerational matrix, organising the data by generation to look for patterns and themes, setting the data against the backdrop of the recent increasing public, policy and research interest in dementia to try and interweave biographical, generational and historical timescapes.

Qualitative research has faced criticism for lack of clarity regarding the relationship between theory and data, and this, I argue, is an important area to address as we continue to develop the contours of QL research. My own perspective has been influenced by Mills (1959), who describes a ‘shuttle back and forth’ between theory and data. I have utilised such an iterative approach, and have drawn on theory from the sociology of chronic illness and family and relationship sociology to develop understandings of the intergenerational experience of young onset dementia.

References:

Alzheimer’s Society (2015). Dementia 2015: Aiming higher to transform lives. London: Alzheimer’s Society.

Henderson, S., Holland, J., McGrellis, S., Sharpe, S., & Thomson, R. (2012). Storying qualitative longitudinal research: sequence, voice and motif. Qualitative Research, 12(1), 16-34.

Mills, C.W. (1959). The sociological imagination. London: Penguin.

Ribbens McCarthy, J., Holland, J., & Gillies, V. (2003). Multiple perspectives on the ‘family’ lives of young people: methodological and theoretical issues in case study research. International Journal of Social Research Methodology, 6(1), 1-23.

Sabat, S.R., Johnson, A., Swarbrick, C., & Keady, J. (2011). The ‘demented other’ or simply ‘a person’? Extending the philosophical discourse of Naue and Kroll through the situated self.  Nursing Philosophy, 12(4), 282-292.

Saldaña, J. (2003). Longitudinal qualitative research: analyzing change through time. California: Alta Mira Press.

Thomson, R. (2010). Creating family case histories: subjects, selves and family dynamics. In Thomson, R. (Ed.) Intensity and insight: qualitative longitudinal methods as a route to the psycho-social. Timescapes Working Paper Series No.3.

Guest post #4, Libby Bishop: Data from the past and for the future – Qualitative longitudinal data available at the UK Data Service

We are pleased to have Dr Libby Bishop contribute as a guest blogger. Libby Bishop (Ph.D.) is Manager for Producer Relations at the UK Data Archive (University of Essex).  She provides support and training on data management toLibby Bishop researchers and data producers, with specialisation in ethics of data use: consent, confidentiality, anonymization and secure access to data.  She also teaches workshops on secondary analysis of qualitative data. Libby worked as a Senior Research Archivist at the University of Leeds, where she was responsible for creating and managing the Timescapes archives and providing support for those using the data. Libby has published individually, and with others, on data management, qualitative secondary analysis and the ethical issues associated with big data. Her work has been critical in supporting the sharing and re-use of qualitative data, and advancing a more nuanced understanding of secondary data analysis.

In this post, Libby explains how a data archive is the perfect starting point for those new to qualitative longitudinal research.  

Data from the past and for the future: Qualitative longitudinal data available at the UK Data Service

You may already be a member of the tribe of qualitative longitudinal (QL) researchers if you are reading this.  But what if you are just starting out? You might be curious about how others have done QL projects.  Of course, there are published articles to look at, and there are many to choose from now. But wouldn’t it be helpful to actually look at the data other researchers have used? To read in some detail what strategies were used to maintain contact between interviews? To read transcripts to discover, for example, exactly how the interviewer gently guided the respondent back to topics from the previous contact, without losing the thread of more recent events? All this, and more, is possible by looking at qualitative longitudinal data collections available for research at the UK Data Service.Time-Background-Clock-Public-Domain

Below I provide a brief introduction to just a few of these collections, of which we have dozens. These are available to be downloaded and used by researchers (after having registered with the Service). Two of these studies are about older age, and another is on a timely issue: elections.

SN 851919 Maintaining Dignity in Later Life: A Longitudinal Qualitative Study of Older People’s Experiences of Supportive Care

The aim of this study was to examine preparations for the end of life made by older people with supportive care needs and the factors that support or undermine a sense of dignity. Thirty-four participants in Bristol and Nottingham were recruited via GPs and day centres. All had health problems that required support and care to varying degrees, including family care and support, medical treatment, community nursing, home care services and moves to care homes. They were interviewed face-to-face on four occasions (on average) between June 2008 and January 2011 and contacted by telephone between interviews. Face-to-face interviews were recorded and transcribed verbatim.

SN 5237 Adding Quality to Quantity: Quality of Life in Older Age, 2000-2002

The broad aim of the study was to define the constituents of quality of life in older age. The research questions were twofold: how do older people define and prioritise quality of life, and how do they feel it can be improved? This study represented a unique multidisciplinary and mixed methods collaboration between investigators with backgrounds in sociology, psychology, social gerontology, transport planning and clinical epidemiology. Following the fielding of the questionnaire, 80 respondents were selected for an in-depth interview to probe factors further affecting quality of life.

SN 6861 Qualitative Election Study of Britain, 2010

This research project recorded the views and concerns of Britons before and after the 2010 General Election. By conducting 14 focus groups with people in England, Scotland and Wales the project investigated, qualitatively, pre- and post-election views. The aim was to generate data that: 1) provided insights into the views and perceptions of citizens on politicians, party leaders, and political issues (e.g. civic duty, political alienation, political activism) before and after the general election; 2) allowed for analysis of the meaning that underlies their assessments, uncover sources of normative values, and make explicit the tacit assumptions participants use to reach their judgements. Three additional focus groups were conducted on the night of the first ever Leaders’ Debates and the transcripts record people’s expectations in advance of the debates and their reactions afterwards. As well as the focus group transcripts, the collection includes a quantitative file of results from the pre-focus group questionnaire given to participants.

And watch this space – comparable data from the 2015 UK elections will be arriving shortly.

What could be better than QL data? Getting funded to do research with QL data! The ESRC has a programme, the Secondary Data Analysis Initiative, which does just that. It offers funding for up to 18 months and £200,000 for research that collects no new data, but uses data from selected existing resources. One of the designated resources is Timescapes, a rich lode of QL data.  Another, the Qualitative Archives on Ageism and Conflict, is held at the Northern Ireland Qualitative Archive.

As always, if you want any help getting starting or looking for data, just get in touch.

Guest post #3, Prof Rachel Thomson: Case histories in QLR

rachelthomsonRachel Thomson, Professor of Childhood & Youth Studies at the University of Sussex, writes our third guest post. Rachel is also one of the directors of the Sussex Humanities Lab. She has been involved in several qualitative longitudinal studies and has co-edited two special issues of the International Journal of Social Research Methodology on the topic in 2003 (6:3) and 2015 (18:3). She blogs at http://blogs.sussex.ac.uk/everydaychildhoods/ and https://newfrontiersqlr.wordpress.com/

In this post Rachel explores case histories in QLR. We too have been contemplating different understandings of ‘the case’ and Rachel’s post has encouraged us to re-think what we consider to constitute a case and how we might go about analysing cases across an archived dataset that comprises six QLR studies. We found her thoughts on the fluidity of cases and practices of casing very insightful. Rachel and colleagues will be running a two-day training course on ‘Case Histories in Qualitative Longitudinal Research’ at the University of Sussex on 6-7th October 2016. We welcome your comments and ideas. 

Case histories in qualitative longitudinal research – some thoughts

Questions of scale

Qualitative longitudinal research can play around with our ideas of scale. A study can seem to be ‘small’, following 6 cases for example, yet at the same time can be ’big’, or perhaps a better word is ‘deep’ in collecting many instances of data for that case over an extended or just intensive period of time. Discussing this point Lyne Yates (2003) makes a case for QLR having a different kind of ‘warrant’ – or relationship with validity –moving us away from ideas of cases as being ‘representative’ in an abstract way –be that they are typical or that they may provide insight into a wider population through the operations of probability sampling. By following just cases over a period of 10 years (as we have in an ongoing study) we are able to understand relationships, sequences, consequences and antecedents in a concrete way- exploring the relationship between what we say and do, and between what we want and what we get – as researchers and as participants. More recently Liz Stanley has challenged the qualitative quantitative distinction on her work using collections of letters showing that in an era of digital data qualitative material and quantity and quantification are not mutually exclusive (Stanley 2015). Rather we might think of scale in terms of a zooming in and zooming out of perspective, and the potential to combine the affordances of the microscope and macroscope. Debates about scale within QLR parallel debates about scale within ‘big data’ and the kinds of digital tools that can be used to explore patterns, to zoom in for the close-up and to zoom out for the landscape or the map.

From cases to casing

QLR can be designed in different ways in order to reveal different kinds of cases. At the most basic level we might think of the case as a unit of analysis that we follow over time. So for example in our project Making Modern Mothers, the unit of analysis was women about to have their first baby. Yet cases are not stable, especially when pregnant and in this study we expanded our case to include significant others (especially grandmothers) and children when they were born. These children are now the focus of a follow-on project that explores digital childhoods, yet the backstory of the family is a vital part of the case and family members play a key part in narrating the case of the child who is the focus and who as we watch moves from being a ‘case’ of a child into a teenager and an adult. Analytically we can also think of the case in other ways, for example thinking about all of the urban families together and considering their affinities and their difference from the rural families. We might also think of the case of social class, or cutting the data set in the opposite direction, from the diachronic to the synchronic, considering how the families responded to a key external event such as the ‘credit crunch’ that turned into the extended period of austerity through which we continue to live. Rather than thinking of cases as stable and defined simply through existence we might follow Charles Ragin (1992) to think about practices of ‘casing’ in social enquiry, a flexible analytic practice that pays due respect to the complexity of the social realm and which in linking ideas and evidence had the potential for the testing and emergence of theory.

The case history and the archive

The ‘case’ itself is an object and genre with a history linked to practices of natural history, collecting, sorting and narrating and reflections. Butterflies were collected and displayed in a case long ago in a way that has parallels with the ways that doctors and lawyers began to conceptualise case histories and case law. A special issue ‘On the case’ of the journal Critical Inquiry helps us as social scientists understand our practices in historical and cultural context as well as helping us see the kinds of spillages that echoes that may travel between medical, legal, scientific and literary uses (Berlant 2007).There is no definitive way of constructing or telling a case, yet we may find ourselves being drawn into particular tropes taking up associated forms of authority. When telling the story of an individual over time it may be hard to escape the perspective of the doctor or the social worker who is able to see and describe underlying causes or pathologies. Perhaps we need to deliberately disrupt these well-worn narrative tendencies by reading materials against the grain, changing the direction of our analysis, or moving between individual and collective or conceptual cases self-consciously in order to find new perspectives.

In earlier work I suggested that we might make use of the notion of the archive more fully in our work learning from the critical work that has been done of reading the archive (Thomson 2007, McLeod & Thomson 2009, Thomson 2011). If we think of our data sets as archives, which can be organised into all sort of cases (individuals, institutional, geographical, temporal), we can also think about the kinds of stories that can be told from the archive, putting material together in a particular way will enable a particular history. Yet this is not definitive or exclusive. That material could be told in different ways by different analysts without taking away from the ‘validity’ of the material itself. Digital information systems allow individuals to build their own archives, copying and linking data from public collections and potentially making their own archives available to others. Sociological data sets are also made available to and interrogated by secondary analysts and there is a compelling case for social scientists to build on the lessons of historical and literary scholars about archival methodologies and epistemologies as well as understanding the new methodologies of the digital humanities. Having my data used by secondary analysts encourages me to believe that the potentials of this area are just beginning to be explored by sociologists – see for example http://www.whiteswritingwhiteness.ed.ac.uk/blog/archive-project-sendoff/

An event

The more I think about the case, case studies and case histories, the more I feel that they lie on the emergent boundaries of new kinds of sociology – even though the case study is associated with the very birth of the discipline and the Chicago School. On October 6th and 7th my Australian colleague Professor Julie McLeod and I will be teaching a 2 day advance methods course for the NCRM, based in the Mass observation Archive at the University of Sussex . The course responds to the request for more focus on methods of analysing QLR data among the participants at our 2015 course ‘Affective methods: capturing everyday temporalities with QLR’. In this course we will be exploring what it might mean to make a case from data archives, both those we generate from primary research, those we find in archives, and those we construct from a range of heterogeneous sources. The course will explore methodologies of the boundaries of history and sociology and between scientific and humanities paradigms. Please join us. http://www.ncrm.ac.uk/training/show.php?article=6361

References

Berlant, L. (2007) ‘On the case’ Critical Inquiry: special issue 33 (4)

McLeod, J. & Thomson (2009) Researching Social Change: qualitative approaches, Sage.

Ragin, C. (1992) ‘”Casing” and the process of social inquiry’ in Ragin & Becker (eds) What is a case: exploring the foundations of social inquiry, Cambridge University Press.

Stanley, L. (2015) ‘Operationalizing a QLR project on social change and whiteness in South Africa, 1770s – 1970s” International Journal of Social Research Methodology 18:3, 251-65.

Thomson, R. (2007) ‘The qualitative longitudinal case history: practical, methodological and ethical reflections’ Social Policy and Society, 6(4): 571-582.

Thomson, R. (2011) Unfolding lives. Youth, gender and change. Policy Press

Thomson, R., Hadfield, L., Kehily, M.J. and Sharpe, S. (2010) ‘Family fortunes: an intergenerational perspective on recession’ 21st Century Society 5 (2): 149-157

Thomson, R Thomson, R. (2014) Generational research: Between historical and sociological imaginations, International Journal of Social Research Methods, 17 (2) 147-156

Yates, L.(2003) ‘Interpretative claims and methodological warrant in small number qualitative longitudinal research’, International Journal of Social Research Methodology 6 (3): 223-32.

 

 

Guest post #2: Dr Fiona Shirani: Visual approaches in QLR

Our second guest post is written by Dr Fiona Shirani, a Research Associate at Cardiff University. She will be leading a qualitative longitudinal work package as part of the interdisciplinary FLEXIS programme, which seeks to investigate how flexible energy systems can meet modern-day energy challenges.

In this post, Fiona reflects on the intersection between QLR and visual methods, reflecting on both her current research and the qualitative longitudinal “Men as Fathers” project, part of the UK-wide Timescapes network. This project explored the transition to fatherhood in both the immediate (over the first year) and longer term (8 years post-birth). One of the many questions Fiona’s work has provoked for us as a team is how to incorporate visual data into our dataset, and the reasons why visuals (and their meaning) can feel more ‘distant’ in secondary analysis than the written word. We welcome your comments and ideas. 

Time to be engaging? Multimodal methods in QLR

Both qualitative longitudinal research (hereafter QLR) and visual methods have seen a surge in interest in recent years, yet relatively little attention has been given to the intersection of the two. This is perhaps surprising given many of the arguments in support of visual or multimodal methods take on a particular resonance in the context of QLR. In this short blog, I draw on my experience of designing and undertaking visual activities in two QLR projects (Energy Biographies and Timescapes) to highlight some relevant issues.

Methodological innovation is an important element of research, and visual approaches have been key to enhancing creativity in qualitative work. QLR provides greater scope for methodological innovation and experimentation due to the extended timescales and flexible nature of the approach. For example, an activity that would take up too much time for a one-off study represents a smaller proportion of a QLR project. There is also time and space for reflection between interviews, giving the researcher an opportunity to hone and adapt activities for later waves of data collection.

Beyond methodological innovation, an advantage of incorporating a range of activities is the potential to make the research experience more engaging for participants. QLR requires a significant commitment from participants and maintaining the sample over time is an important concern. Whilst some participants enjoy the format of a qualitative interview, others may relish the opportunity to direct conversation through a photo-elicitation exercise, for example. Activities can also be conducted between interviews, serving as opportunities to both maintain contact and collect further data.

In both Timescapes and Energy Biographies we used visual activities to encourage participants to talk across extended time frames, thinking about their past memories and anticipated futures. Thinking temporally is a particular concern of QLR research yet can prove challenging for participants. Having a tangible reference point in the form of a visual representation can help anchor discussions.

Alongside these benefits, consideration must be given to some of the challenges arising from using multimodal approaches in QLR. Most notably, whilst accumulation of information about the individual is a strength of QLR, it also raises issues around anonymity and confidentiality. Adding visual data further complicates this issue and requires careful thought about how visual artefacts produced during the research should be analysed and presented, as well as challenges for archiving and data re-use.

In our Timescapes and Energy Biographies projects we have primarily used images as a means of eliciting talk; therefore analysis has focused on pictures and their accompanying text. We have found that narratives often go beyond what is represented in the image and therefore have argued the importance of attending to both talk and text. However, there are many possibilities for the analysis and creative presentation of multimodal data, and participant-generated images also inspired some elements of our public engagement exhibition. We also explored the possibility of working with images alone during a multimodal workshop with academic colleagues, where feedback indicated that the everyday nature of the images made them accessible for people to imbue with their own interpretations. However, there are clearly important ethical issues to consider in presenting images without the contextual information of their production and asking others to engage creatively with them.

Visual or multimodal approaches have much to offer QLR and combining the two could provide multiple benefits to future research.

See our paper in the International Journal of Social Research Methodology for more detail.

Guest post #1, Dr Anna Tarrant: Reflections from the Men, Poverty and Lifetimes of Care study

This is our first guest post, written by Dr Anna Tarrant. Anna is currently working as a Leverhulme Trust Early Career Research Fellow at the University of Leeds. She is leading a research project called ‘Men, poverty and lifetimes of care’ that qualitatively explores how men negotiate their care responsibilities in low-income localities. As part of the study she will be conducting qualitative secondary analysis of two datasets from the Timescapes archive. Susie and I recently met Anna to chat about our respective projects. While their aims are different, they share many methodological and conceptual issues. How, for instance, should we define a case? How does our epistemological positioning shape our approach to QL data analysis? Should qualitative secondary data analysts forge relationships with the primary researchers? How can we keep a strong purchase on time and temporality when looking across different projects? And can software help us? We hope Anna will be kind enough to blog again at a later stage in her project to reflect on some of these questions. In the meantime, you can follow Anna’s own wonderful blog here.

Assessing the feasibility of secondary analysis within and across two qualitative longitudinal datasets; reflections from the MPLC studyAnna image

Like the NCRM project that Susie and Emma are leading, the Leverhulme Trust funded ‘Men, Poverty and Lifetimes of Care’ (MPLC) study has facilitated important opportunities for reflection on key methodological questions about the feasibility of working with multiple qualitative longitudinal datasets. Qualitative secondary analysis (or the re-use of qualitative data in it’s simplest form) is a relatively novel approach in the context of the much wider spread re-use of quantitative data, yet it has already provoked a great deal of debate, particularly in relation to issues of epistemology, ethics and context (see Irwin, 2013). An emerging area of concern within these debates focuses on the possibilities and pitfalls associated with bringing multiple datasets into analytic conversation and whether or not this is possible or even desirable.

In the first year of the MPLC study, I conducted a qualitative secondary analysis on two datasets from the Timescapes archive, allowing me to reflect on some of these questions. At the outset of the proposed study, which aimed to explore men’s care responsibilities in low-income localities, I identified the Following Young Fathers (FYF) and Intergenerational Exchange (IGE) studies (see the Timescapes website for more information about both studies) as possible resources for exploring key themes in relation to this substantive area of focus. For the purposes of rigor, I employed a three stage methodological strategy that was attentive to the principles of the ‘stakeholder ethics’ model (Neale, 2013) including:

  1. Familiarisation with the datasets (by having individual conversations with available members of the original research teams and reading project outputs),
  2. Holding a data sharing workshop to:
    1. Consolidate the familiarisation process and,
    2. To facilitate a more collaborative mode of working with the original project teams by bringing the datasets into analytic conversation with a focus on the broad themes of men and care,
  3. The Qualitative Secondary Analysis itself.

These processes are discussed in much greater depth in a Timescapes Working Paper.

QSA across multiple datasets has been really insightful and productive. While a time-consuming and difficult process (particularly in becoming familiar with data generated by others), it has fed directly into the design and conduct of the second empirical phase of the MPLC study, for which, I am interviewing men living in low-income localities. In combination, the FYF and IGE datasets have also provided a sampling framework for the MPLC study. Since September 2015, I have focused on recruiting men living in low-income circumstances of different age groups, in order to explore men’s trajectories and their care responsibilities over time. I have also been able to recognise the importance of men’s wider interdependencies in low-income localities and this has prompted me to ask the participants in the MPLC study, specific questions about the significance of their wider support networks. In terms of substantive outcomes, I have gained greater insight into men’s experiences of living on a low-income over time and how gendered inequalities mediate these processes. While the datasets are not directly comparable, the participants in both studies live in contemporaneous times and there are remarkable similarities across the datasets with regards to how men experience low-income life. In bringing the datasets into conversation, it has been possible to test my emerging theories with empirical data from both datasets.

For more information about the MPLC study please follow the study on Twitter @menpovcare and the study website http://menandcare.org.uk

Welcome to the project

Welcome to the ‘Working across qualitative longitudinal studies’ collection of guest blog posts contributed by experts in the field. We created this collection of blog posts for two reasons. First, we were conscious that accounts of data management and analysis in qualitative research are often sanitised by the time they reach academic journals. We were, therefore, keen to document and share the trials and tribulations and decision-making processes underlying such analysis, thereby contributing to debates around good practice. We also wanted to kick-start conversations about analysis/secondary analysis across large scale and/or multiple qualitative data sets. The guest posts range from early career researchers through to international experts, and address topics as varied as the ethics of using big qual data, using secondary qualitative data and computer-assisted qualitative data analysis software. They profile the diversity of QLR and big qual that is taking place internationally.

The blog collection gathers the experiences and perspectives of those conducting analysis across large-scale, multiple and/or qualitative longitudinal data sets, particularly the re-use of archived material. The possibilities comprise multiple permutations from bringing together two or more archived data sets through to combining archived material with a researcher’s own primary data. Some of the blogs focus on secondary analysis and the re-use of archived data sets, including QLR material, whilst others are concerned with handling large volumes of qualitative data. In both cases the approach undertaken involves, to some degree, engagement with an amount of data that otherwise would be challenging for a single researcher or small team to handle with qualitative integrity