Advanced Statistics [MSc] postgraduate personal statement example

Personal statement example

On Saturday mornings I help organise the volunteer rota at a food bank near my flat, and for a while I also kept the spreadsheet recording what came in and went out. I noticed that our shortages of tinned fish and long-life milk were not random: they clustered in the weeks after school holidays. Showing this with a simple plot changed how we requested donations, and it reminded me that much of what interests me about statistics is the gap between noticing a pattern and being confident it is real. I want to study advanced statistics at postgraduate level because I have reached the point where my undergraduate tools let me see that gap clearly but not always close it.

My degree in Economics and Statistics gave me a solid grounding in probability, linear models, inference and econometrics, and I graduated with a 2:1. The modules I found most rewarding were those on generalised linear models and on Bayesian methods, where I first saw how prior assumptions can be made explicit rather than hidden. For my dissertation I used anonymised, publicly released data on missed GP appointments to model the probability of non-attendance using logistic regression, with predictors such as appointment lead time, weekday and practice-level deprivation. The most useful part of the project was dealing with its limitations. Practices sharing a local area were clearly not independent, so I compared a standard model with one including random intercepts for practice. The fixed effects barely moved, but the standard errors widened noticeably, which taught me how easily a tidy result can overstate its own certainty. I would like to understand hierarchical models properly, including the theory behind estimation, rather than relying on software defaults.

Since graduating I have worked as a data assistant at a regional housing association. The role is not senior: I clean repair-request records, produce monthly reports and answer questions from managers. Even so, it has been valuable preparation. Repair data are messy, with duplicated jobs, free-text categories and dates entered in inconsistent formats, and I have become careful and patient about checking data before analysing it. I wrote R scripts to automate a report that previously took a colleague most of a day, and I documented them so others could rerun them. When a manager asked whether a new contractor had reduced the time to complete repairs, I realised that a straightforward comparison of averages ignored the many jobs still open at the end of each month. Reading about survival analysis to handle that censoring is one of the main reasons I want further training; I could recognise the problem but lacked the depth to treat it rigorously.

Alongside work I have kept studying. I worked through much of An Introduction to Statistical Learning, completing the R labs, and I found its treatment of the bias-variance trade-off a helpful counterweight to my habit of adding predictors. I have also been revising multivariable calculus and matrix algebra, because I know a postgraduate course will demand more mathematical maturity than my applied modules did.

Outside data, I play five-a-side football each week and have organised our team's fixtures and kit money for two seasons, which is mostly a lesson in chasing people politely. The food bank work matters to me for its own sake, though it has also made me comfortable explaining numbers to people who are busy and sceptical.

I am drawn to a demanding taught programme because I want both stronger theory, particularly in likelihood, Bayesian computation and mixed models, and the discipline of applying it to real problems. In the longer term I hope to work as a statistician in public services such as health or housing, where careful analysis of imperfect administrative data can inform decisions. I am ready for the step up in rigour, and I would bring practical experience of messy data and a habit of asking how sure we really are.