Computational social science postgraduate personal statement example

Personal statement example

My sociology degree included a compulsory second-year module on survey design, and I expected it to be the part I endured rather than enjoyed. Instead it was the module that reorganised how I thought about the rest of the subject. We were given the Understanding Society questionnaire and asked to work out why particular questions were worded as they were. Seeing how much interpretive care sits behind a single variable made me want to work with data rather than around it, and it is why I now want to study computational social science formally.

My final-year dissertation grew out of that interest. My local council ran an online consultation on changes to bus routes, and the responses were published as a long PDF of free-text comments. I combined a descriptive analysis of secondary travel-to-work data with a simple content analysis of around four hundred of those comments, coding them by hand first and then using Python to check whether term frequencies matched the categories I had created. They mostly did, but the interesting cases were the disagreements: comments about "reliability" turned out to cover both timetable accuracy and personal safety while waiting, which my keyword approach flattened into one thing. I wrote that limitation up honestly and argued that automated text methods need hand-coded validation samples rather than replacing them. My supervisor encouraged me to submit it for the department's undergraduate research showcase, and it was selected for a short presentation, which was a modest but genuinely encouraging outcome.

Statistically my training is solid rather than advanced. I have worked through regression, including logistic models and some multilevel modelling, in both R and Stata, and I understand the difference between a coefficient I can interpret and one I have simply produced. My programming is more self-taught: I learned Python largely through the dissertation and through following Allen Downey's Think Stats, which suited me because it approaches distributions and hypothesis testing computationally rather than through formulae alone. I have also read Salganik's Bit by Bit, and his distinction between data that are found rather than designed has stayed with me, particularly the point that large digital datasets are often not representative in ways that are hard to detect. I am aware that agent-based modelling and network analysis are areas where I have read introductions but written almost nothing, and I would want the MSc to push me there.

Outside study, I have worked weekends at a garden centre for three years, mainly on tills and the plant advice desk. It has taught me to explain things to people who have not asked for a technical answer, which I think matters more in this field than it is usually given credit for. I am also membership secretary of a community running club of about ninety members. When I took over, attendance was recorded on paper registers that nobody consulted. I moved them into a structured spreadsheet and produced quarterly summaries; these showed that our Thursday beginners' session had far better retention than the Saturday one, and the committee shifted our coaching effort accordingly. It was a small piece of analysis, but it convinced me that the useful part is usually framing a question the group actually has.

After the MSc I would like to work in applied social research, ideally in local government or the third sector, where quantitative and computational methods are still often kept separate from the qualitative work that gives them meaning. I want the training to be able to do both properly, and to be the person who explains clearly what a model can and cannot support.

Review this personal statement

Latest reviews