Thursday, 17 May 2012

The loveliest statistician (i.e. geek)

The other day I had an exam in the morning, it was two hours long and a bit stressful. After joining the other half for lunch (amazing chilli nachos), I spent the afternoon with a good friend. We spent about four hours working on her medical dissertation project, which needed some statistics. I have the statistics programme, SPSS, went to lots of statistics courses last year (which I STILL have yet to be reimbursed for) and my dissertation last year was basically all statistics, which I had to teach myself and do myself, so I can sort of hold my own when it comes to medical statistics.

Although it took a fair bit of brain-effort to recall everything I’d forgotten over the past year, this came in very useful for my friend. Her project has gone from having some percentages that look quite nice, to having Chi-squared for trend with a p value less than 0.001 – to anyone who knows research, that’s gone from “somebody found that the intervention probably changed things a bit” to “they found that changes resulting from the intervention were highly significant, with a less than one in 1000 possibility that they were just caused by chance, therefore these interventions are good and everyone should adopt them”. We also added in some nice standard deviations because ranges should be left behind at Maths GCSE.

Surprisingly successful for a post-exam afternoon, where the temptation to just slump is strong. Instead we spent hours fiddling with SPSS, transferring her data from Excel to SPSS, making new data files, fiddling until we found some results that meant something, were the right type of statistics, and were reproducible for the rest of her data. But the worst of it is: I really enjoyed it. I feel I shouldn’t admit to this. This is statistics after all – possibly the most boring thing I’ve ever had to learn about throughout both my medical and genetics degrees. It is the realm of geeks of the dullest kind. I’m not going to deny my geekness, I did do a genetics degree after all, but statistics geekery is a few steps too far; I am not that boring! Learning about stats is mindnumbing, but actually implementing it, playing with SPSS sot that you get something shiny and publishable – I’ll be honest, that is fun. It’s experimentation, with useful results.

I feel I should embrace my inner stats-geek, and have a potential idea: I’ll become a FREElance statistician, as a junior doctor. Here’s the deal – you (consultant, researcher, clinician, nurse specialist, whoever) have a research project that you hope will show something exciting, and you’ll want to tell the world about it. I am an academic junior doctor, and I will do your boring statistics for you. Give me your data, and I’ll give you the shiny publishable results. If you ask nicely, I’ll even write your results up for you (I like a nicely displayed table). I will not write anything else – your introduction and discussion and all the wordy bits are yours alone (I hate writing). I’ll write the paragraph on how I analysed the results, I’d have to really. And the best bit: I’ll do all this for free. All I require is that you put my name on your publication or conference abstract, after all, I will have contributed to it.

And if you’re really organised, I can discuss your data collection methods with you before you start, to ensure you’re collecting the best data to get the best results.

Do we have a deal? What’s not to like?! I would love to know if this would actually work in reality. If you (non-existent readers) know anyone who needs some medical stats… Trying To Be A Medical Student is here to help – you know where to find me.

Win-win situation: the researchers who don’t have time or understanding or patience or SPSS get their statistics done for them, I get more research experience, and I get published, which can only help my academic career. And will add something non-gynaecology-related to my CV, and since that’s not an area I plan to go into (never going to be a surgeon), that would be useful.

Here’s to embracing the inner geek. And to helping friends out.

No comments:

Post a Comment