Biostatistics postgraduate personal statement example

Personal statement example

In the pharmacy where I work as a dispenser, we keep a small tray for prescriptions people never collect. After a few months I began noting which medicines ended up there, mostly out of curiosity, and noticed that repeat prescriptions for statins appeared far more often than antibiotics. It was an informal tally, not a study, and I could think of half a dozen reasons for the pattern that had nothing to do with patients' choices. Working out which explanations my notes could and could not rule out is the kind of question I want to spend the next stage of my education learning to answer properly, which is why I am applying for postgraduate study in biostatistics.

My degree was in biology, but the modules I enjoyed most were the two in experimental design and data analysis. They taught me to fit linear models in R, check residuals, and think about what a confidence interval actually says rather than treating it as a decoration on a bar chart. For my dissertation I grew tomato seedlings under three watering regimes and measured their height every three days for six weeks. My first analysis treated every measurement as independent, and my supervisor pointed out that repeated readings from the same plant are not new information in the way separate plants are. Reading up on mixed-effects models, mainly through the lme4 documentation and the relevant chapters of Faraway's Extending the Linear Model with R, I refitted the data with a random intercept for each plant and tray. The treatment effect remained, but the uncertainty around it widened noticeably, and I had to rewrite my conclusions more cautiously. That change felt more satisfying than the original result, because I understood why it was more honest.

I also spent time on the parts of the project nobody grades. I wrote a short script to flag implausible readings, such as a seedling shrinking by four centimetres overnight, which usually turned out to be my handwriting. I kept a plain text log of every decision about excluding or correcting data. These habits seem small, but in clinical and public health data, where measurements come from many hands, I expect them to matter a great deal.

Dispensing has given me a practical sense of how health data is created. Every label I print and every record I update becomes, somewhere, a row in a dataset. I have seen how a dose recorded as "one twice daily" can be entered three different ways, and how a missed collection might mean a patient has stopped treatment, moved away or simply used another branch. I am not a clinician and do not want to overstate what I see from behind the counter, but it has made me wary of assuming that a variable means what its name suggests.

Outside work I keep score for my netball club's second team, which has turned into an ongoing argument with myself about whether shooting percentages across a season of uneven opponents tell us anything at all. I also share an allotment with my sister, where my attempts at a controlled comparison of two potato varieties have so far been defeated by slugs that do not respect randomisation.

I am aware of gaps in my preparation. My mathematics stops short of a full course in probability theory, so over the past year I have worked through an introductory probability textbook alongside the free linear algebra lectures from MIT OpenCourseWare, doing the exercises rather than just watching. I would welcome rigorous training in survival analysis, longitudinal methods and the design of clinical studies, and I hope eventually to work as an analyst in public health or clinical research, where careful statistical reasoning shapes decisions about real people. I bring a biologist's familiarity with messy data, steady working habits from a busy dispensary, and a genuine eagerness to learn the theory properly.