1/07/2010. Contributed by Geoff Willmetts
pub: Wiley-Blackwell. 238 page indexed small enlarged paperback. Price: £27.50 (UK). ISBN: 978-1-4443-3436-4.
check out website: www.wiley.com/wiley-blackwell
All right, 'How To Read A Paper' it might seem a rather odd book to choose for SFCrowsnest, but with SF we're also interested in real science as well, not to mention there are many science professionals into SF. In that respect, this might sound like a busman's holiday but when offered a selection from Wiley's catalogue, the interest came from how to read medical reports. You see enough of them quoted in the newspapers but how many of you look up the medical journals they are sourced from, let alone understand what is said? I came away from this book thinking the name is a misnomer and it could easily have been called 'How To Write A Paper' and wished the titled was a bit cleared by calling it a 'Medical Paper', lest someone thinks it means a newspaper.
The biggest surprise I got from author Trisha Greenhalgh's book, in its fourth edition since 1995, was how little there is for an understanding of statistics by those writing medical reports. That and where wildcard information turns up that is outside of the pattern accepted, isn't checked to ensure that there is something wrong with the data or patient than just treat it as an error. When it comes to pharmaceutical companies also adjusting reports to favour their products, this becomes a minefield. A very dangerous minefield, especially where human lives are being looked after.
What Greenhalgh points out is how to look at the information, which is really the same way you write up a physics or chemistry experiments at school. Look at the objective, the test carried out - even if it's done with multiple test subjects and the study of the results. At any stage, there is a need to ensure that the correct information comes out. With drugs having so many different side-effects, mostly because humans aren't consistent or average, you would have thought that pin-pointing the type of person the drug would not be effective with would be safe practice. Yes, it is probably done and I suspect it can't be tested against all drugs and their combinations but if you're on medication, it can look a bit disconcerting that the list of side-effects out-numbers the paragraphs of what it's supposed to treat. Then again, the test subjects are chosen where the particular problem the drug is supposed to help alleviate don't tend to be representative of the people out there who might have other problems as well. No wonder doctors cross-check medications to see if they are incompatible. I suspect this is, in part, because so many doctors read Greenhalgh's book and better practice has developed since her first edition. The fourth edition actually takes into account new changes since then so it isn't just a new printing but a revised one.
Something else comes out of this from my own observation is that school essays these days are based largely on what can be gleaned off the Internet even if teachers do try to get their students to put things in their own words. Imagine this technique carries on into adulthood and scientific papers. It does make you wonder if teachers should be asking their pupils a bit more as to what it means and get them to draw conclusions to get their brain-matter working. This might not always get the standard replies or interpretations but that shouldn't be a penalty neither because at least they'll be questioning things which serve them in good stead when they move onto university and jobs.
What is even more worrying from Chapter 12 is getting statistics first and drawing a conclusion from something discovered in the pattern as medical practice. I should point out Greenhalgh doesn't like this any more than me, especially as the scientific approach is to have a hypothesis first and see if there is sufficient evidence to back it up. If there isn't but presents something different, then you revise the hypothesis and point out the original theory could not be substantiated. This is one chapter you should really read and question any statistics you read that your medical problems fall into and you seem outside of. Statistics can mislead, especially if they are draw out the wrong result simply by getting the wrong sampling. Getting them wrong trivialises statistics and makes people disbelieve them so there is a good argument for better practice not to mention repeat sampling to see if it's consistent. With four editions, Greenhalgh does point out that there are some changes in medical testing now but I also think the samplers and those who are asked for information need to consider more what they are being asked rather ramble out an answer. I've done some phone surveys in recent months and refuse to answer ambiguous badly worded questions because it asks for a loaded answer rather than what I should actually say.
For such a small book, this will give you a lot to ponder on. I need to remind you people this is more an academic book than for general use. I do think it or a similar version was out there for the general public and at a lower price so it's open to all. Greenhalgh is an excellent writer and puts things over in a clear way. I should point out that you shouldn't ignore statistics or medical articles but the book does show you how to put your brain into thinking gear and not just absorb information without thinking about it making this an excellent book. If I can react so strongly in this review, then it has to be worth something.
Add SFcrowsnest.com daily news updates to your own web site or blog - just cut and paste the code below...
Stephen Hunt's novels - USA