Readers of my book – and of these blogs – will know that I make greedy use of news items about research findings in the press. But in 2007, it is apparent that there has emerged a degree of scepticism about some of the messages which researchers convey, especially with respect to health matters. One item I particularly liked was in the Times (21.12.07) which disposed of eight medical myths – ideas that had assumed scientific proportions but for which no empirical evidence could be found. In particular, the notion that drinking eight glasses of water a day is good for you appears to be based on nothing more than fantasy – or rather on a misreading of the evidence.  Apparently, the paper that the idea is based on suggested that 2.5 litres a day is a suitable water allowance for adults but contained the sentence, “Most of this quantity is contained in prepared foods”. Only today, I saw a mother and her two children taking the dog for its half-mile walk and all were carrying water bottles. Empirically, it seems, such maternal concern is misplaced.  

When I began these blogs a year ago, it was with some reluctance, I have to confess. I am not a natural blog-writer: I don’t live a very interesting life and I am not especially opinionated. I am, at heart, an empiricist – constantly fascinated by the way that scientists set about the task of answering questions, solving problems and contradicting unproven assumptions. My editor at Palgrave Macmillan said, “Don’t worry about it – but just remember who your audience is”. I have tried to do that

I wanted, at the start of this new year, to say something about qualitative methods – and I can do that best by drawing on my situation as [what you would think of as] an elderly man. As I have aged, I have become more and more intrigued by the failure of the media to reflect the fact that a third of the population, or more, is aged 60-plus, and I’ve come to realise that that is because the people who write for the papers or work in television are all in their 20s, 30s or 40s. Columnists write about chatting up people, being ditched, having babies, finding the right schools, teenage problems, divorce and so on; if they mention old people at all, it is usually about the role of the younger generation vis-a-vis their ageing parents . 

To some extent, I have found that the same phenomenon affects the choice of research topics among students. But one of the great benefits of qualitative methods is that it can enable researchers to go outside their own sphere of experience and explore the experiences of other people entirely. That, after all, is what anthropologists traditionally have done, and they are the source of much qualitative methods thinking. Qualitative methods studies could aim to find out about the ordinary reality of old age – instead of perceiving it only from the writers’ own perspective.

Yet again a couple of survey items in the news media grabbed my attention this month.

First, an outrageous claim quoted in one paper: “The humble custard cream has topped a poll to discover the nation’s favourite biscuit. Britons placed it above the bourbon and the shortbread when asked to name their teatime treat of choice. The [custard cream] took 93 per cent of the overall vote. Next in line . . . is the bourbon with just over 2 per cent. Further down were the cookie and the ginger biscuit, garnering the support of just 1.5% and 1.19% of respondents. The digestive received less than 1% of the vote. More than 7000 people took part in the poll, organised by a maker of wheat and gluten-free products.”

Do you believe it? I don’t. 93% against 2%! That kind of statistical outcome sounds more like a Stalinist election result than the product of a well-run survey seeking an objectively reliable result.

Were participants offered just those choices? Does the company that organised it have a vested interest in custard creams? And notice the pseudo-scientific use of two decimal places for the ginger biscuit.

Next, there was a superb one-page piece on epistemology in The Times on the 3^rd November (page 35). If I was editing a book on research interpretation, I would certainly want to include this feature by the paper’s health editor, Nigel Hawkes – though quite what the Times would charge me for the privilege I shudder to think.

The paper is prompted by the understandable confusion in the public mind concerning the logic that flows from research experiments in respect of diets and other aspects of ordinary life – especially, at the time, focusing on the alleged dangers of eating bacon. Hawkes spells out how difficult it is to organise experiments – and how it’s even more difficult to interpret the results. He spells out the differences between placebo-controlled trials, observational studies, cohort studies, case control studies and meta-analyses. He tries to explain why readers should always treat with caution news stories that appear to tell us the causes of various health conditions – but the problem is that the headline writers tend not to encourage caution, because they want to grab the interest of the readership – and they do that by preying on our anxieties.

And there I was: all set to comment on this month’s news item that pinpointed suburban wine drinkers as being the biggest group at risk of alcohol-related ill-health. The basis of the report was the usual one of men and women drinking in excess of 21 or 14 units per week respectively. And I was going to say that I had always wondered where the ‘recommended limits’ came from and what research techniques had been used in their calculation because, as I note in my book, asking people retrospectively about what they have drunk provides you with a notoriously risky piece of information. And, given the assumption that in respect of alcohol consumption, it is done, not just retrospectively by a week, but, in the case of people, for example, suffering from cirrhosis of the liver, over a lifetime, then the reliability of the data must have been highly questionable.

But then The Times upstaged me. The journalist, Andrew Norfolk, found that the figures 21 and 14 “were set in stone” 20 years ago after the Royal College of Physicians produced a report on alcohol misuse. “Yet”, reports

Norfolk

, “these guidelines have no basis in science. Rather, in the words of a member of the Committee that drew them up, they were simply ‘plucked out of the air’.” A former editor of the BMJ, Richard Smith, told the The Times that the figures were not based on any clear evidence. The epidemiologist on the Committee, David Barker, was quite clear that “it’s impossible to say what’s safe and what isn’t”.

But there was a strong feeling among the Committee members that they had to come up with something in order to guide the medical profession. And so, says Smith, “Those limits were really plucked out of the air. They weren’t really based on any firm evidence at all. It was a sort of intelligent guess . . “

The government nonetheless accepted the figures as sound and they have formed the basis of policy-making and practice ever since. But, says, Norfolk, a host of epidemiological studies have, since then, produced evidence to the contrary: in 2000 a WHO report proclaimed ‘low levels of risk’ for men as being less than 35 units were week; and in 1993, an Oxford-based study of 12,000 male doctors found that the lowest mortality rates – lower than teetotallers – were among those drinking between 20 and 30 units a week.

All of which illustrates how careful researchers have to be when it comes to the interpretation of statistics.

My academic website and email pages are full of advice about how to detect plagiarism or cheating by students.  There are even conferences held to discuss the issue and training provided by universities to help lecturers deal with it.

I hate the idea.  But I understand how it comes about.  Of course, the internet has aggravated the situation.  Even website pages that draw users’ attention to books also contain sponsored ads that offer access to an infinite range of student essays on all subjects.  The implication is that you can just download and present it as your own.

In theory, asking students to do research projects should reduce the scope for cheating – but, of course, it doesn’t.   Even professional researchers have been found out fabricating evidence in order to prove a theory – and in the social sciences, it is extremely difficult to guard against blatant dishonesty.   A friend of mine at a university in Greater Manchester was appalled to discover that his research assistant had been sitting at home just filling in the questionnaires that he was supposed to have been putting to 150 parents of young children.    My friend had to sack his assistant and then start all over again from scratch.

The trouble is that, as in life generally, you have to trust people to be reliable. Most people are, but not all.  The big opinion poll companies have ways of detecting when their interviewers cheat – simply by seeing whether the responses they are getting fit within the pattern of normal distribution overall. But in the small scale academic world of private research, it’s almost impossible to be certain . . .

I was caught out once by a graduate student who had written what I thought was a really excellent dissertation on a rather unusual topic.   I was editing a major book at the time, and his/her piece of writing was ideal for one of my chapters.    I did some work on it to improve the style and flow, and then asked the student for their permission to use it – in return for a very modest payment.   I was very surprised not to get a quick reply – normally students are over the moon if you suggest that their work is fit for publication.    Eventually, after numerous phone calls and several letters, I finally managed to get the go-ahead.   It duly appeared – but 18 months later I happened to be reading a book on the same topic as the student had focused on, and, to my horror, I realised for the first time that much of what had been written had been plagiarised from this book.  Fortunately (!) the author of the book was no longer living, but I felt total consternation at something which I felt – indeed, was - responsible for.     Needless to say, the chapter was omitted from the next edition. No wonder my student had been so reluctant to give me permission for publication!

I recount this story to show how terribly easy cheating is – and how difficult it is to counteract.   My hope is that everybody who does a research project using my book as a guide will do so without recourse to such underhand methods.    If you do cheat, you may – or may not – get a good mark, but you won’t have learnt anything in the process.

In the UK, August has always been the holiday month. The schools are empty.  In places like Great Yarmouth – down the road from where I live - it usually rains a lot, and those who have opted for family holidays at home suspect that it is costing them more than if they’d gone to Ibiza or Turkey.   

But for teenagers, it’s also the time when A-level grades are published – with triumphant or devastating consequences for students who’ve spent two years striving to get good results.  It’s hard to exaggerate their importance.  In my case, I squeaked a pass in French (I’d always been hopeless at languages – and I still am), and this, alongside good grades in History and English meant that I could go to Liverpool University to do Social Science.  (I also got into LSE, but my Auntie in London warned me off it because, she said, it was full of communists.  It wasn’t, of course, and it wouldn’t have mattered if it was; but I was young, naïve and impressionable – and at that age you look for guidance from any ‘reliable’ quarter.) 
The social science course in Liverpool was then wonderfully multi-disciplinary, and although it meant that I could never afterwards call myself a pure psychologist, sociologist, historian or political scientist, it gave me a rounded, critical, non-ideological perspective on the world which has stood me in good stead ever since.   It also gave me my first taste of research methodology – though, in truth, the statistics course was terrifyingly unsatisfactory.  I only succumbed to the allure of research and became professionally skilled at it when at the age of 27 I took work in the civil service as a Research Officer.   Research methodology is a discipline that you can only learn by doing it; as with playing the piano or painting pictures, merely listening to lectures is pretty useless.

For those about to embark on university student life, good luck!  And if you have to do a research-based project at some point, I hope you find the book useful.   Let me know whether it proved helpful – or how I could make it better.