I’ve just returned from Hyderabad where I spent four days meeting our advisors, discussing the Connectors Study with them and thinking about how we can best work together.
Over the next four years we will be working closely with colleagues in Athens and Hyderabad to carry out the study and so issues around cross-cultural research have been at the forefront of my mind in the run up to Connectors ‘going live’.
Most challenging of these issues so far, and still largely a work-in-progress, is sampling.
Since Connectors started a couple of months ago the most common response we’ve encountered to the study is ‘yes, that’s all very nice, but which children?’ A fair question and one we are taking our time answering.
Over the next four years we will be following children from age 6 until age 10 in our three study cities. So one thing we do know about these children is that they would have been born in 2008 and have spent their early childhood growing up in post-crisis, austerity politics. Beyond that we have some work to do.
In the original proposal I wrote about sampling across levels of affluence which is fairly typical and uncontroversial (I think!?) but feels slightly unimaginative. Conversations over the last month with colleagues in Athens and Hyderabad, as well as at Sussex, are helping us to think about the sampling strategy in more creative ways.
For example, Rachel Thomson and colleagues’ work carried out in the methodological innovation network in Qualitative Longitudinal Research (QLR) suggest that for qualitative longitudinal research it is important to strategically select cases that are emblematic of larger trends (Thomson 2009). Suitably selected cases can help the QL researcher to analyse their phenomenon in a way that cuts across global and local situations, macro trends and micro issues, while remaining mindful of the temporal. A sort of dynamic less is more approach to sampling that is targeted from the outset.
But what informs that targeting? Taking the lead from emergent QLR approaches and other recent writing on sampling in qualitative research, there are a few things that we are considering.
We have started looking at what might qualify as a ‘trend’, or trends, in each country and city. For example, a recurring news story from Greece post-2008 has been the stark rise in mental health problems and in particular the rise in suicide rates amongst men. The rise of unemployment, and in particular youth unemployment is another trend.
Meanwhile, recent research carried out by the LSE’s Centre for the Analysis of Social Exclusion tells a tale of two cities in London and, amongst other things, the decentralisation of poverty from the inner city to the outskirts of the capital.
The theme of cities within cities is also salient in Hyderabad. During my visit our advisors spoke about four Hyderabads: the Old City, the New City, Cybercity (Cyberabad) and Secunderabad.
A big trend in Hyderabad over the last 10 years has been the IT boom, with many technology companies now located there and in Bangalore. We were introduced to the income variations that have been made even more stark by these developments and unevenly distributed across the four cities within the City. For example, the affluence of Banjara and Jubilee Hills and the poverty of the slum areas dotted around the city, as well as different characters of each slum depending on their location and dominant population.
We were also advised about the continued importance of caste in Indian society, and we heard about mobility in and out of caste categories through conversions to Christianity or inter-caste marriage, an old but continuing trend.
Alongside looking at trends we are also spending a good deal of time imagining our findings so that we can work backwards to identify appropriate sites of emergence.
According to Nick Emmel, who’s written a great book summarising different sampling approaches in qualitative research, this sort of theoretical or purposive sampling strategy will have us spending a considerable amount of time thinking about where the phenomenon we are most interested in is likely to occur and trying to sample there (Emmel 2013, p. 79).
Our phenomenon is ‘the development of an orientation towards social action in childhood’ and so another question we find ourselves asking is what forms of social action are happening on the ground and what if any role are children playing in them?
Ultimately, we need a sampling logic that will provide us with rich enough data to respond to our research questions in interesting ways, will make sense across three different cultural context enabling us to have sensible conversations across those contexts, and which will allow for manageable heterogeneity necessary for theory development.
One thing that struck me the most about the conversations in Hyderabad last week is how creative and productive it was being in a completely unknown (to me) context, and how it gave me fresh eyes and energy to think about this crucial part of our study. The benefits of working cross-culturally have already started to emerge and I look forward to much more to come.
(And we will let you know what we eventually settle on as a sampling strategy!)