To ‘be’ in the 21st century – let alone get employment, scrutinise commercial offers or function as a citizen – demands data. The ability, that is, to analyse blocks of information, check charts and treat samples with friendly suspicion. There’s the basic, or civic case for improving our capacity to deal with data and since much data is numerical, that implies quantitative upskilling.
Tackling that society wide is a big project, starting deep in the thickets of the primary school curriculum, leading on to the balance of subjects at secondary school and increasing the (by international standards) small proportion of students who do any maths after the age of 16.
Because that project is so big and long term, let’s do what we can, now, in our own backyard. So for the social sciences here’s a bold proposition. An integral part of education in the study of human behaviour is learning how to measure, as part of understanding the methods by which social science finds things out, which make it – let’s not be preachy – ‘science’.
The curriculum for sociology, political studies, geography and so on has to accommodate, and maybe it’s time to look again at what’s taught in those subjects, notably economics and psychology, where quantitative techniques are more central. Of course that also means understanding the social construction of numbers, and research methods. Of course it doesn’t exclude learning how to extract meaning from texts or institutional behaviour. It does inescapably mean getting to grips with elementary statistics, probability, error and correlation.
You can make the case in functional terms. Professor John MacInnes of the University of Edinburgh has been advising the Economic and Social Research Council, and has amassed evidence on the employment benefits for social science graduates of quantitative understanding.
Put at its bluntest, students will earn more and have a wider choice of work if they can show they have acquired some of the skills traditionally associated with the STEM subjects, but have studied a social science subject. They may even have an advantage over their STEM contemporaries.
Understanding data matters
You can also make the case from within the disciplines. Can a sociologist understand the dynamics of society, ethnicity, gender, and economic differentiation without numerical tools; ditto a geographer the contours of place? No. That’s not because the only good data are quantitative, but because the credibility of sociological understanding rests, in part, on handling quantities.
Here’s a contemporary example. The Cameron government says enthusiastically that it believes in evidence – in founding both policy and service delivery on knowledge, testable propositions. Whether ministers are sincere, whether their policies are indeed ‘evidence based’ isn’t the issue. It’s that policy and political engagement depends more than ever on understanding how data is collected and analysed – seeing when randomized control trials are appropriate and where the subject matter simply will not support that particular form of confirmation.
If social scientists are to participate, as citizens or as experts, they have to have the capacity to address and assess evidence. Do ministers’ contentions about families, welfare claimants, tax credit recipients stack up against the large amounts of empirical – especially longitudinal data – that have been building up? Answering the question requires a skill set that is too often missing from the social science undergraduate curriculum.
The Q Step initiative
That’s why the ESRC – with the higher education funding councils and the Nuffield Foundation – have been pushing the case for quantitative methods to become a requirement. The latest initiative sees 15 universities signing up to a ‘step change’ in quantitative social science skills. Of course the extra money in the initiative is an incentive.
But with it must come a change in consciousness and conviction that accepts the study of social science will – must – get ‘harder’, as it foregrounds methods, statistics and data analysis.
The prize, however, is glittering. It’s more than the rejuvenation of disciplines such as sociology, it’s all-round improvement in civic conversation, as a larger proportion of the graduate population bring to bear better understanding of opinion polling, of the claims made by advertisers and the way the media (and ministers) so often substitute for the whole population an outlier, an extreme example.
Many people, not just on the left of politics, have been dismayed by the tenor of recent debates about welfare, above and beyond necessary concern about costs and affordability in an ageing society. Those debates have, often, rested on quantitative propositions, advanced as fact, that even small inquiry and limited understanding of numbers would call into question.
Better quantitative education for social scientists does not need to be justified as a ‘progressive’ move. But it is.
David Walker was till the summer chair of the ESRC methods and infrastructure committee and a member of council. He is a visiting professorial fellow at the Institute of Education.