Sunday, October 7, 2018

Unlike the USA, most academic sociologists in the UK and Australia have no expertise in quantitative methods or statistics

From Academic Sociology and Social Policy Think Tanks in Britain and Australia: A Personal Reflection by Peter Saunders. Not ground shaking or uncommon but interesting in the details and the comparisons.
After 24 years teaching sociology at the University of Sussex, I made a double career-switch in 1999, moving from Britain to Australia, and from university employment to working for public policy think-tanks. Reflecting on my experiences of British and Australian sociology, as well as the academic and applied policy worlds, what stands out most is the ideological conformity and closure of academic sociology in both countries.

1.2 Part of the problem is the lack of respect and understanding for empirical evidence. Unlike the USA (where I taught in 1996), most academic sociologists in the UK and Australia have no expertise in quantitative methods or statistics. They justify their ignorance through fatuous appeals to anti-positivist philosophy, which in practice means their claims are rarely testable and are not even expected to be so. This has enabled left-wing and feminist ideas to maintain a stranglehold on the discipline, and few sociologists in either country see this as a problem. Despite the image it has of itself as 'open', 'tolerant' and 'critical', academic sociology in Britain and Australia is, in my experience, none of these things.

1.3 Like most young people drawn to a career in sociology, I started out on the left. My politics were reflected in my early research and, I suspect, in my teaching. Writing on class, housing and the state at a time when Althusserian Marxism was driving the urban sociology agenda, I made a reputation for myself as a 'left Weberian,' challenging Marxist orthodoxies, but from a safely-socialist standpoint. My career blossomed. Articles were published, publications got cited, invitations were received to prestigious foreign conferences, research grant applications were favourably reviewed, and promotions followed.

1.4 But during the 1980s, my politics changed. After falling out with the Labour Party over its refusal to allow working class people to buy their council houses, I started to wonder why they should not also be free to choose the schooling and health services they wanted. Before I knew it, I was attending lunches at the Institute of Economic Affairs and reading Friedrich Hayek. This was not a good career move in a discipline where the gatekeepers were (and still are) overwhelmingly socialist (90% of sociology professors in Britain describe their politics as moderate or far left: Halsey 2004). While friends and colleagues at Sussex remained friendly and collegial, the sociological establishment outside the university became increasingly antagonistic. Our growing estrangement can be traced in my CV.

1.5 In the mid-1980s, I was on four journal editorial boards. By the early 1990s, I was on none. For a period in the mid-eighties, I was serving as external examiner at three different universities simultaneously. A few years later, all the invitations had dried up. It was the same story with PhD examining, and with peer-reviewed research grants. The quarter of a million pounds of ESRC funding I received in the 1980s had dwindled by the 1990s to a single, £8,000 personal grant (a sum small enough not to require approval by peer review).


As an academic in Britain, I was censored by the process of peer review. As a government researcher in Australia, I was censored by bureaucrats and politicians. But writing for think-tanks in both countries, I am not censored at all. I write what I want, and if they sympathise with what I am saying, they publish it. If one think-tank doesn't like it, another will take it, for I now operate within a genuine free market in ideas.

1.16 In my experience, the pressure to intellectual conformity is much higher in academic sociology than in policy think tanks. It's just that most academics do not realise it, for they spend their lives swimming with the ideological currents rather than making any effort to go against them.
It is worth noting that American sociologists having superior expertise in quantitative methods or statistics doesn't fundamentally improve the quality of their ideologically flawed work. In Britain and Australia, they don't have the statistical chops so their work is not framed in a fashion to be tested (the scientific method), they simply make bald assertions. It cannot be refuted because it cannot be tested.

In the US, sociologists, psychologists, etc. have a higher propensity to use quantitative methods and statistics but the quality of their work remains nearly as low. They go through the motions of the scientific method and do correlations and the like but while they are superior to their British and Australian brethren, as 'scientists' working with ideologically tinged assumptions, their work overwhelmingly fails to replicate. Yes, they use maths, but their maths is wrong.

The problem is not so much their maths as their ideological assumptions which undermine their work. They are willing to learn some quantitative methods but they are unwilling to give up their ideological assumptions.

This is illustrated in the Implicit Association Test debacle. The underpinning ideological postmodernist intersectionalist assumption is that all white Americans are inherently racist (and misogynist) and that this inherent misogyny and racism infects all institutions as well as their everyday behavior.

Deployed with fanfare from Harvard twenty years ago, it has gained wide acceptance in victimhood circles and been used as the basis for performance reviews, for remedial coaching, for public policy, etc. Actions driving hundreds of millions of dollars in public expenditure and incalculable impact on personal careers.

And all for nought. While it looks rigorous and formal and scientific, it is not. People taking the test at different times end up with different scores (i.e. it is not measuring any stable and consistent attitude). There is no observable correlation between scores and behaviors. African-Americans are as prone to bad IAT scores as white Americans. It is all a Potemkin Village of scientific rigor over a foundation of ideological racial grievance.

Despite their trying to put the best possible spin on every weakness and failure of IAT, even the Wikipedia article acknowledges:
The IAT is the subject of much controversy regarding precisely what it measures, and the lack of reproducibility of many of its results.
After twenty years of embarrassing failure, I sense that the tide has turned within academia, and while still extensively used and studied, the more credible members of the field have begun to abandon IAT.

Not so in civic life. Government agencies and schools and even corporations still pump hundreds of millions of dollars into activities based on IAT scores. They are waving an animist wand at a vestigial social problem in order to gain social kudos for appearing to 'try'. Very much comporting with the ethos of participation awards.

No comments:

Post a Comment