Reading between the lines on Research – Part One

Research can produce some spectacular headlines. I was looking at a newspaper this weekend and read that ‘23% of women hospitalised for anorexia met the diagnostic criteria for autism’. On the surface, this could mean that nearly a quarter of women with severe anorexia are autistic. That is a hefty piece of news. But, because I was reading a very balanced newspaper it went on to report that more work needs to be done and this study only included 60 women. We don’t know from this information how all the other variables were taken into account (for example; age of women affected, length of time affected, education, social class etc.) And in fact only about 14 women showed a tendency. A tiny illustration of how careful we all must be when we read about research.

Research is a loose term covering everything from a double blind randomised controlled trial (RCT *) to a quick poll of people in your gym at 08.00 on a Monday morning. There are some key points that are useful to consider when you read about the results of a piece of research.

I like to know who funded the research. Did the funding body have a vested interest in the outcome? Do Coca Cola really want us to find out that fizzy drinks are harmful to us? But it is the big giants that have the money for these large-scale projects.

I also want to know what questions were asked (look in the appendices if you have access to the research paper). You know the saying ‘Lies, damn lies and statistics’ – research can be carefully targeted to get the results the funders want. So, look at the whole picture, use your own knowledge. If you wanted to know if fizzy drinks cause weight gain, what would you want to know? Did the researcher ask all those questions? It is hard and our own prejudices get in the way. If you want to find out how hard it is, give it a try – I bet you’ll find yourself asking the questions that will elicit the answers you want to get.

Look at when the research was carried out. Is this new and exciting or has this been known for a long time and possibly superseded by something more up to date? Attitudes and opinions change and we want to know current thinking wherever possible.

Consider too how many people were in the study and how long the study lasted. Asking 5 people a set of questions as a ‘one off’ will get different results to closely observing a cohort of 500 over 10 years. Both are valid but you, as the reader, need to be aware of the limitations of the study.

This leads onto the next point – if this study was done again, would it get the same results? If someone else asked the same questions would they get the same results? If it was an experiment in a lab was it sufficiently well controlled to be replicated accurately? Be sceptical, look for accurate, unbiased studies.

Its all about controlling the variables, more about this next time in part two. (*I’ll also explain what I mean by a double blind RCT!)


This blog was written by Marion Foreman

Marion is a nurse with many years’ experience and a personal trainer. Her two specialist areas are training older people and training people on their cancer journey. Marion sometimes works with very elderly people in care homes, helping them to remain mobile and independent.

Marion and her husband own two gyms and are passionate about helping people to be their best. Marion is always keen to make sure that their members are given up to date and relevant advice about living a healthy life style.

Dont be shy, tell us what you think!!