Last month, we published my article about how a newly-qualified data scientist views his prospects (Incoming data talent: A student perspective). It received some feedback on Twitter. As well as hearing about the curriculum, one reader said it would also be beneficial to know about the cost to do a data science Masters course and where a prospective student could obtain funding.
So I decided to apply some of my own newly-acquired skills, having recently done an online course called "Doing journalism with data". I thought I would to try to visualise some information about data science masters courses.
The course told me what I needed to do to scrape data. It seemed simple enough. I put in a formula at the top of a Google sheet and it would automatically populate the spreadsheet with all the data that I wanted from a specific web page. It didn’t work. I figured that scraping pages of search results probably doesn’t work in the same way as scraping a one-page list.
So I got my data the manual way. I went through pages and pages of search results from UCAS and and filled up my spreadsheet. The data included the course name and price, the university, as well as the city and country. Super basic, I know, but I wanted to start small with 100 records looking at courses across England, Northern Ireland, Scotland and Wales.
On average, a Masters in data science costs £8,617
The average cost of a full-time, on-campus Masters course with a strong data science component of the courses I found was £8,617.05.
I found out that the most economical and the priciest courses are both in London. Data science at Middlesex University is a pocket-friendly £4,500 while a course of the same name at the London School of Economics is £26,976 – a six-fold difference between the two.
Cool story, bro. But not really. I didn’t feel that information was worth making an infographic about. Looking back, I should have collected more data. People reading this would probably want to know which universities offer scholarships and bursaries, what they cover, other sources of funding.
However, worked with what I got and produced this graph. It’s basic and imperfect, but one has to start somewhere. Onwards to the next viz! (That’s how us data pros refer to visualisations - I know the lingo already.)