I've been accepted into a combined B.A. and M.A. program focused mainly on political science, international relations, and the history of the SWANA region. I'm starting in October and have some time to prepare for the coursework, including a future thesis.
I already have some experience with Python, Pandas, and Matplotlib, and I'd like to become more proficient at using them for research and data analysis. I'm also considering learning SQL and the basics of Git, since both seem useful for academic work and future employment.
Should I stay focused on Python, or would SQL be especially valuable for political science and research? How deeply should I learn each tool, and what would be a practical way to build these skills if I'm not pursuing software development?
3 Answers
Keep building on Python, since you already have a foundation and it’s useful for cleaning data, analysis, visualization, and automating repetitive research tasks. The best way to improve is to work on projects you actually care about: analyze election results, public opinion surveys, historical datasets, conflict data, or government statistics.
SQL is worth learning at least to a basic or intermediate level. You don’t need to become a database engineer, but being able to filter, join, group, and aggregate large structured datasets is very useful in data analysis. Treat it as a practical research skill rather than something you need to master immediately.
Don’t overlook Git, but don’t treat it like another huge programming language either. You can learn the essential workflow in an afternoon: initialize or clone a repository, check changes, add files, commit, review history, and restore earlier versions.
Git is useful for code, research notes, Markdown files, writing, and other projects because it gives you a history of your work and makes mistakes easier to undo. Once the basics feel natural, you can learn branches and remote repositories. Using it regularly on small personal projects is more valuable than studying every advanced command.
So Git could also be useful as a general version-control and knowledge-management tool, not just for programming? I work on writing, design, CAD, and 3D projects too, so that sounds useful.
A good path would be to strengthen Python fundamentals, then learn the research-focused parts of the ecosystem: NumPy, Pandas, Matplotlib or Seaborn, Jupyter, and eventually tools for statistical modeling. Also spend time on data cleaning, documentation, reproducible workflows, and explaining results clearly. Those skills often matter more than knowing a large number of libraries.
For academic work, try to reproduce a published analysis or create a small project from a publicly available dataset. That will teach you how to find messy data, document assumptions, handle missing values, and communicate conclusions. SQL can be learned alongside this when you start working with larger relational datasets; the basics are quick to pick up, but advanced SQL can become important in professional data-analysis roles.
That’s what I had in mind: using programming mainly as a way to make research and analysis easier rather than becoming a software developer. I enjoy Python because of what it lets me do, so I’ll focus on projects related to my academic interests.