Health data science postgraduate personal statement example

Personal statement example

My final-year project began with a fairly narrow question: whether a simple regression model could describe variation in emergency admission rates across English local authorities using openly published NHS and census data. What I found interesting was not the model itself, which explained less than I had hoped, but how much of the work sat before the modelling. Reconciling boundary changes between data years, deciding whether to treat suppressed small counts as missing or as bounded values, and documenting each choice so my supervisor could follow my reasoning took more time than fitting anything. That experience convinced me that health data science is largely a discipline of careful decisions about imperfect data, and that I want formal training in it rather than continuing to improvise.

My degree gave me a solid grounding in probability, linear models, and statistical inference, and I chose optional modules in computational statistics and survey methods. I was comfortable with the mathematics but taught myself most of my practical computing. I moved from spreadsheets to R during my second year, largely because repeating a cleaning process by hand three times was enough to persuade me, and later learned enough SQL to query relational databases without assistance. I read Hadley Wickham's R for Data Science while doing this and found its emphasis on tidy, reproducible pipelines genuinely changed how I organise work: I now write scripts that regenerate every figure from raw files, which saved me considerable trouble when I discovered an error in one of my project's input datasets a fortnight before submission.

Since graduating I have worked as a data administrator for a small charity providing community health and wellbeing services. My role is unglamorous and useful. I maintain the referral database, check incoming records for duplicates and inconsistent coding, and produce quarterly activity summaries for trustees and funders. I have learned how much depends on how frontline staff actually record information under time pressure, and how easily a well-designed field becomes meaningless when it is optional. Working with a colleague from the service delivery team, I revised our intake form and its guidance notes so that ethnicity and postcode were captured more consistently; completeness for those fields improved noticeably over the following two quarters, though I would not claim the change was the only reason. I also began producing the summaries from an R script rather than assembling them manually, which reduced the preparation time from most of a week to about a day.

Outside work I keep a personal project going with the openly published English prescribing data, looking at how prescribing of a few common medicines varies between practices and how much of that variation persists once list size and deprivation are accounted for. It is deliberately modest in scope and mostly a way to practise handling large files, joining datasets with awkward identifiers, and presenting results honestly, including the parts I cannot explain. Away from screens, I help organise a monthly board games evening at my local library, which involves rather more emailing and chair-moving than strategy.

I am applying for postgraduate study because my current skills have clear limits. I want proper training in epidemiological study design, causal inference, machine learning methods suited to clinical and administrative records, and the legal and ethical frameworks governing access to patient data. In the longer term I would like to work as an analyst within a health system or research unit, producing evidence that service planners can actually use. I am prepared for the workload and looking forward to being taught by people who know where the pitfalls are.

Review this personal statement

Latest reviews