UK government's Year of Code gets off to a bad start

The Year of Code is supposed to teach all the UK's school children to code, but even if you support this fashionable American idea, this doesn't seem to be a useful way to do it
Written by Jack Schofield, Contributor

The UK's £500,000 Year of Code project is an attempt to teach the country's 5- to 16-year-olds to code, with the idea that this will help them in their future lives. They will be able to create their own web pages, and a few may even get jobs coding. This is the sort of "motherhood and apple pie" project that the IT industry can unthinkingly support. However, judging by the YearOfCode website, it looks half-baked.

Almost every company that hires staff -- not just IT specialists -- would like them to be more computerate, but is this the way to get them? Not really.

First, it's not clear there is any educational value in teaching children to code, especially not the average 5 to 11-year-old. You'd think an educational project for use in schools would have educationalists involved, and perhaps the support of companies that specialise is this market. But most of the YearOfCode's advisors appear to be the founders of small start-up companies such as Songkick, and one is "Managing Director, Strategy and M&A, British Gas". I'm sure they are all excellent people, but they're not Seymour Papert, Mitch Resnick or Alan Kay.

UK bloggers Tom Morris and Adrian Short have both pointed out the woeful unsuitability of the YearOfCode's advisors, one of whom has already quit. Emma Mulqueeny of Young Rewired State found "advisors" didn't actually do any advising and concluded: "I want nothing to do with this."

Second, the website doesn't offer even the barest details of a viable strategy. Worse, what the project's executive director Lottie Dexter revealed in a BBC Newsnight interview (starting at 5'32") left me with the feeling that it is non-viable. Teachers can't code, and the idea that they will learn enough in one day to teach children to code beggars belief. It doesn't make sense to implement this approach unless it has been thoroughly tested, and there's no evidence of testing on the YearOfCode website.

Lottie Dexter on BBC Newsnight.
Lottie Dexter on BBC Newsnight. Source: ZDNet screen capture

Lottie Dexter admits she doesn't know how to code, and she has no discernible qualifications -- such as educational or work experience -- for the job. However, her Newsnight interview neglected to mention that YearOfCode is an independent charity, isn't setting the curriculum, and isn't producing material for teachers. Apparently that's been done by the British Computer Society, which also got over £2 million to set up a network of 400 Master Teachers to train classroom teachers. In other words, YearOfCode is little more than a PR exercise, which is why it's being run by someone who is basically a PR.

Still, Dexter claims "you can build a website in an hour, completely from scratch". And she's right. Indeed, you can do it in less than an hour if you use a simple template. Whether this teaches you any computer technology skills is another matter. I don't believe it does. At that sort of level, the best it can do is teach you how to use some parts of a mark-up language, which is (being generous) only just coding. It certainly isn't programming. It's not even remotely close to "computer science", of which coding is but a small part.

Third, there's nothing new here: YearOfCode is just recycling some currently fashionable but equally fatuous American techno-romanticism. In the UK, it's a bit of Tory party flag-waving that is meant to make them look like they are doing something about the "skills crisis" when they are not.

What all this has to do with the new computing curriculum, which will be compulsory in UK schools from September, remains a mystery.

Days of future past

It's not as though the UK education system hadn't already made some serious attempts to teach computer literacy. In the 1980s, for example, the government subsidized the sale of Acorn BBC Micros (and other computers) to schools, and the BBC ran an associated computer literacy project that included well-produced TV programmes.

At the time, the coding was in BBC Basic, and some children did learn something, though this expensive project's long-term impact isn't obvious at the moment.

Many UK schools also introduced children to programming concepts using floor turtles and the Logo language. This is far more likely to help develop real computing insights than learning a few HTML tags, though research has shown that even this approach doesn't ensure children understand what they are doing. See: On the cognitive effects of learning computer programming (PDF).

I'm all in favour of trying to teach children technology skills, including coding, but that should be based on what we -- and educational researchers at MIT etc -- have learned over the past 40 years. Useful tools include things like Logo, Scratch, and Pivot's Stickfigure Animator animation program. This also gets round the problem that a great many UK schoolchildren can barely write correct English, and their maths skills are poor by global standards.

Based on the Khan Academy and similar websites, perhaps the YearOfCode's £500,000 would be better spent on creating some YouTube videos that teach children some computer science in an entertaining way. One series that springs to mind is Unplugged, from New Zealand, which manages to do that without even using computers. The short episode on binary numbers, for example, would probably expand Lottie's understanding, and still leave her 53 minutes to finish her website.

Editorial standards