Medical statistics postgraduate personal statement example

Personal statement example

My interest in medical statistics grew out of a fairly ordinary frustration. In my second year I took a course on regression and was given a dataset on patient recovery times with a substantial number of incomplete records. The tidy answer was to drop them; the tutor pointed out, almost in passing, that in a real study those patients might be the ones doing worst. That comment stayed with me, because it turned a technical decision into a question about who ends up represented in a conclusion. I have been reading and working towards medical statistics ever since, and I would now like to train formally in it.

My degree in Mathematics and Statistics gave me the foundations I will need: probability, linear models, likelihood-based inference and computational statistics. For my final-year project I used a publicly available dataset from a completed cancer trial to compare Cox proportional hazards models with parametric survival models, focusing on what happens when the proportional hazards assumption is questionable. I spent a long time on the diagnostics, and learned to be careful about how I described the results. The honest conclusion was modest: the models agreed on the direction of the treatment effect, but the estimated benefit at five years varied enough that I would not have wanted to quote a single figure without explaining the assumptions behind it. Writing that up clearly taught me more than the modelling did.

Alongside this I have built up practical skills independently. I learned R properly in my first year because the teaching used a different package, and I now use it for everything, including a small personal project simulating dropout in a two-arm trial to see how much bias different handling strategies introduced. I have also worked through parts of a Bayesian modelling textbook and written a few simple Stan models, though I am aware I am at the stage of following worked examples rather than designing my own approach, which is one reason I want structured teaching in this area.

I have also learned to work with other people on something with real constraints. In my final year, two of us were asked to produce a short statistical summary for a student welfare survey run by the union. My part was cleaning the responses and producing the tables; my collaborator wrote the commentary. The interesting difficulty was that several questions had been worded so that answers were not comparable across years, and we had to decide what we could reasonably report. We ended up presenting less than the committee hoped for, with a note explaining why, and I still think that was the right call.

My part-time job has shaped how I think about communication. As a shift supervisor in a supermarket I spend a lot of time explaining decisions about rotas and stock to people who are busy and not particularly interested in the reasoning, and I have become better at giving the essential point first. I also help run a weekly community walking group, where I keep the sign-in records and sometimes chat to participants about why the local health team asks for attendance figures. It has been a useful reminder that data collection is done by people, with all the gaps and approximations that implies.

What draws me to a master's in medical statistics is the combination of demanding methodology and direct usefulness. I am particularly keen to study clinical trial design, missing data methods and epidemiological study design in depth, and to gain experience of working with real data under proper governance. In the longer term I hope to work as a statistician in a clinical trials unit or a public health team, contributing to studies where careful analysis genuinely changes what is recommended.