Our HE dataset in focus this month is the recently published 2018 entry – March deadline UCAS applicant figures.
Spring has put a coil in our tails and we have left best practice chart types tucked in the winter coat pockets and all gone a little bit curvy! Why? Well we thought we would flex our Tableau muscles and have a go at a few chart types we haven’t yet had a bash at to see if we can make ’em work for this project.
Adam took a liking to a Chord chart, Dave some circles and Elena got sooper curvy with a lovely no data prep Swankey Sankey!
UCAS publish a set of statistical pdfs and csv files full of all sorts of delicious figures describing applicants and applications from the 2018 UCAS cycle as at the March 24th deadline. Why? March the 24th is the deadline for some Art and Design courses, and here at #VisusualisingHE towers, we didn’t want to forget these creative types missed in the UCAS Jan deadline data releases (most known in the publication calendar). Hence, we waited especially for this release to capture them as well.
The tables published include applicant numbers by age, sex, country of domicile, ethnic group, POLAR3, POLAR4, SIMD 2016, country of institution applied to, and institution type (higher, medium, and lower tariff), as well as the number of applications (choices) by subject group.
Key takeaways from the main UCAS domicile headlines:
- Applicants by domicile – Non UK applicants up 8% compared to previous cycle, EU (excluding UK) up a lesser amount of 2%, UK down by 3%, overall all domicile down by 2%.
- Applicants by UK domicile – Northern Ireland displaying the largest percent change in applicants (-5%), England (-4%), Wales (-3%), overall UK down – 3% on previous cycle.
- Applicants by English region of domicile – The North East showing greatest volatility with -8% applicants, however the North East do form the smallest population of English domiciled applicants (2018- 15,820).
- Applicants by declared country of domicile – China remain the stand out country of domicile applying to study in the UK, accounting for 20% of all non EU applications (13,070 / 65,440).
Adam has been wanting to have a bash at a chord chart for a while, so headed straight to a fantastic instructional post by Noah Salvaterra called DIY chord diagrams in tableau that has been saved in his favourites for sometime.
In his blog post Noah basically guides you through how to prep your data and helpfully and very generously shows you how to clone his tableau file and replace it with your own source data. From then on it’s up to you to get creative with the look and feel of the viz.
It is fairly rare in HE data that you get a dataset pop up that is perfect for this chart type, given that you need matching dimensions to show a ‘to and from’ (country of domicile of applicant and country of provider). All I had to do was add some blank rows to help scaffold the non UK applicant data (because I didn’t want to exclude them).
Adam settled on vizzing Table 7 contained in the UCAS overview which takes a look at the country of domicile of applicants and the UK country location of the provider.
Interactive viz: UCAS 2018 Entry | Domicile
Dave, well he got creative with circles, and came up with a novel way of presenting the data set, in a FT style. He also opted for the big picture vs. high level of precision, so where Adam and Elena spent time in getting those numbers in the visible space or the trendy #VizInTooltips, Dave kept it simple – no numbers as values are encoded in the relative colour and size of the circles.
Dave’s viz definitely grew on us quite quickly even though it’s kinda a bubble chart… which is generally a big ‘no no’ in the #Dataviz community. But here, it just works because it is simply not trying to say too much.
Dave focused on the country location of both the applicants and the the HE providers. His method elegantly shows whether applicants chose to apply to providers based locally, and if they choose the go elsewhere in the UK, whereabouts they chose to go.
So what are the key insights? The largest proportion of each group of applicants, based on their location, apply to institutions in their own country. Not many students outside of Northern Ireland choose to apply to NI providers but NI applicants don’t necessarily choose to stay locally as they also apply to study in HEPs in England and Scotland, too. It was also interesting that the majority of applicants to Welsh HEPs by volume were actually English.
Interactive Viz: UCAS 2018 Entry | Location of Applicant vs Location of HEI
Elena took on a Sankey, tried it many different ways: old school, hard way… the data prep way… several times… Unfortunately, this didn’t quite work, so she then took a punt at the Information Lab’s no data prep method. After a battle involving an undisclosed number of attempts, a few choice expletives and a delicate navigation of the nested table calcs, she struck gold and mastered the Sankey build in a record time of 7 mins (this still includes following the steps outlined in the blog post religiously)!
One word of advice: Ian Baldwin‘s blog post is fantastic – it is written very clearly and has extremely useful screenshots of what your table calcs should look like. Just remember, when ha syas ‘make sure your calculations look exactly the same’, then make sure they are EXACTLY the same!
So what’s so special about the Sankey diagram? Well, Sankeys can show movement, a flow. It is frequently used to show poll data to show the proportion of voters parties loose or gain between two elections.
In this case, Elena chose to show what proportion of the total applicants chose each subject and how that differs between the gender and country region of the applicants. The best insight is seen when hovering over the right arm of the Sankey on either chart, or on the subject titles in the table at the top. This will reveal how popular this subject was for each group of applicants.
Interactive viz: UCAS 2018 Entry | Gender & Subject Comparison
The minute it was published it got Viz of The Day!
As usual thank you for reading and we hope you enjoyed exploring our vizzez!
Dave, Adam and Elena