Following on from last week’s exercise in turning the 2015 South African matric results into an interactive map, we’ve had a quick go at this year’s performance report and plotted out the per-school results for 2016. These results were made available in the annual performance report, published today on the Department of Education’s website.
A few technical notes – there were 81 schools which featured in last report which weren’t in this one, unless they’ve changed their name, and 21 new schools to take their place.
I’ve included some details that people requested from the last map, including whether schools are in the public or private sector and the 2015 student:teacher ratio. Some of the public/private sector classifications are missing, and I don’t know how reliable the student and teacher number reporting is. For some rural schools figures reported in the official EMIS data suggest ratios of 1500:1, so I’m assuming this is to do with part-time staff not being reported or just poor reporting in general.
There’s still around 400 schools out of the total 6 898 that I haven’t found location data for. These are now on the map – I’ve arbitrarily assigned them a place in the ocean just off of Durban. So if any can provide me with GPS co-ordinates of where they should be, please drop me a mail.
Here’s what last year’s map looked like:
You can immediately see one big difference: a lot of schools in the west of the country which scored less than 50% last year scored higher than 50% this year. I haven’t had time to do a great deal of investigating, but to me it looks like a lot of those schools are very small, so the difference between 45% and 50% can be just one or two pupils. Or class sizes can vary wildly from year to year – Joyi Senior in Mthatha, for example, entered just 55 learners this year compared to 104 last year, so while the same number of learners passed the percentage doubled.
Of course, if you zoom in a bit things do start to look a bit more similar once the red and green dots are separated out.
The overlay below shows a slightly different view. The size of the dot indicates the reported teacher to staff ratio (ignoring schools like Solomon Mahlangu in Nylstroom, which officially has a class sizes of 1 469 students). I mentioned before that there’s no clear link between class size and achievement – although that’s possibly due to poor data or low achievements in small rural schools offsetting the benefits of small class sizes elsewhere.
Breaking that down by quintile – ie. the wealth of schools – is interesting, however. Small schools and low incomes do seem to produce worse results. The graph below is a chart of overall school size (including non-matric years) on the x-axis, and quintile on the y-axis. Colour is red for lower than 50% pass rate. The schools marked in quintile 0.5 are those which don’t have an official designation (I think these are largely new, private establishments).
I haven’t included data on progressed students in this chart, but will put together some graphs on that over the next few days.
I’m chatting to Code4SA about hosting this data for download and interactive visualisation. In the meantime, let me know if you find any errors.