A slow knee jerk (computing student retention)

I'd like to introduce David Evan's Unqualified Remarks blog on the BCS website. It's a recent addition to the blogsphere, and already has a number of interesting, informed and thought-provoking posts. I expect that that I'll be linking to his posts - of course I may not agree all of the time!

One post that I thought was interesting was 'A quick knee-jerk on computing drop-outs'. It speculates from a Computing article that reports on the National Audit Office's report on graduate retention. In short Maths & Computing degrees have the lowest retention and completion rates (I don't like the term 'drop-out' - sounds like blaming the students). His speculations broadly were around three points.

1. ICT as taught in schools is a poor fit to that in universities - more on that later...
   
2. A maths/numerical/science background is probably more relevant to a computing degree than school ICT - no argument there, but perhaps this is worth unpacking further in future posts.
3. That since applications have dropped in recent years (since 2001, by almost half) universities are taking in students with lower A-level grades.

David did remark that he did not have research data to hand. Thankfully, there is some evidence base around computing retention. As head of a computing department you'd imagine I'd be looking at this closely - and you'd be right. There are some studies that I do have to hand that are relevant to David's argument, and I'd like to share them with you.

Let us start with the speculation around whether the drop in applications for computing degrees since 2001 has lead to poorer-quality students getting in.

I have taken a graph from the Developing the Future 2007 report that compares the performance of applicants for computing degrees to the overall pool of applicants (I assume that this is based on HESA data). As we can see there has not been a drop in the performance of computing applicants.


I admit that this is a finding that I find somewhat surprising as anecdotally I have heard that drops in entrance requirements have occurred (and given the drop in applications you would have expected this).

In any case there is a clear difference in ability between the average performance of all university applicants (around the BBB level), and that of computing applicants (around the CCC mark). So David may well be right about lower grades on entry, but it wasn't the recruitment crash that caused it.

What should be of real concern is that computing is losing ground - the gap in performance is getting wider year-on-year!

NB. I should take the opportunity to say that City has not followed the national computing recruitment trends - we are one of the minority of computing departments to have maintained both student numbers and quality since 2001. But it is fair to say that a lot of computing departments in the UK are under pressure. I'll be posting on this soon, I hope.

Now, onto what is taught in schools. I doubt that I could find any computing admissions tutor that is happy with ICT teaching in schools. To be frank, I find the syllabus of the BTEC National (a vocational qualification) more relevant to studying computing at university than A-level ICT since it focuses on technical issues. But back to the evidence base. The perception of students in schools and colleges has been studied, two relevant ones are:

1. Alison Mitchell et al, ‘Computing Science at University: what do pupils in Scottish schools think about it?’. Presentation at the Higher Education Academy Workshop on Progression, Retention & Recruitment, Edinburgh 2005.
2. Gillian Lovegrove and Anna Round, Report on the HEFCE-funded initiative ‘IT Professionals in Education, Increasing the Supply'. Newcastle/London, November 2005 and January 2006.

They concur with David's suspicions. What is clear from these studies is that students find ICT in schools to be unchallenging and a poor indicator of the challenging and rewarding careers available in the IT industry (or what a degree involves). It also seems to put off the more able students, and attract the less able - clearly not what we want. These studies deserve further analysis - there's a lot to discuss there.

As a final note David did mention that the ICT syllabus problem was (apparently) less in Scotland (though the Scottish study does seem to have very similar responses to the English studies of school attitudes). I'm familiar with the Scottish education system (having lived there for five years), but I've never looked at their school teaching of ICT in detail. But if readers can shed light on this claim, I'd be most interested.

PS. Of course, it goes without saying if you can throw any further light on the above issues - please post on my profile.