A ghost in the machine - observing computer science interviews

Upon leaving my previous role at the University of Essex the very generous leaving present was a palm-pilot. Whilst this was a very useful and appropriate gift, it also played to my rather irrational fear of technology. My age is such that I was one of the last school year groups not to receive computing lessons. At University, it was anticipated that, excluding the final year project, all essays would be handwritten. (The exception to this was my year spent attending college in the USA, when I had to master an electric typewriter). For most of my first job at Newcastle University access to a desktop computer was something that was seen as a rare and beautiful thing, whilst at Essex I resisted getting a laptop and using e-mail until more or less forced at gunpoint. I only started using text messaging about three months ago. One of my resolutions when starting the new post at Oxford was to try and conquer this apathy towards IT, and this blog is in part a result of my new-found interest in technological solutions to problems.

Thus my delight in going to sit in on some interviews for Computer science. Now, bearing in mind yesterday's advice on knowing what factors are significant about the subject discipline at Oxford I discover in the 2007 entry prospectus (page 36) that Computer Science requires

"a sound understanding of mathematical ideas is needed...both for potential applications such as scientific computation, and for reasoning rigorously about the specification and behaviour of programs. Practical skills must also be developed.."

In addition the tutors will

"want to see how you tackle problems and respond to new ideas; they will be more interested in the problem solving process.."

The two tutors had allocated thirty minutes to each interview. The candidates had already taken a two and a half hour written test covering mathematical and analytical skills on the previous Sunday afternoon, although the tutors were not aware of the individual scores that the candidates had achieved at the point that they were interviewed. Prior to each interview the tutors discussed material on the application that related to the candidate's interest in computer science, and this formed the first set of questions when the candidate was seated at the interview table. As has been the case in every set of interviews to date, the focus was on teasing out the applicant's enthusiasm and motivation for their subject. Some obviously had extensive programming experience, whilst others talked about their interest in the impact of computers in society, and the pace of technological change. The interviewers then asked more penetrating questions, for example, in the case of students who had experience of programming, they were asked to identify how they had applied that knowledge.

Following on from this initial area of activity, the candidates were presented with a series of graphs which they had to discuss. The questions and graphs were progressively more difficult to assess and interpret, with the interviewers asking the candidate to explain their reasoning and approach to resolving each problem, eventually (for the last couple of graphs) to use notation to explain why they had reached their conclusions. The candidates were then asked to draw a number of graphs (this is why the interview was conducted at a table!) using information provided by the interviewers - again the construction of the graphs became progressively more demanding. All the candidates were given the same exercises to work on.

The final exercise required the applicant to stand at a whiteboard and resolve a mathematical problem using algorithms. Oxford requires Computer science students to have studied A-level mathematics, and the knowledge required to successfully complete this exercise was pitched at that level.

Throughout the interview the interviewers were happy to clarify any points that the candidate found ambiguous, and this was not held against an applicant in the subsequent discussion of their performance when the interviewers were marking the interview. With each of the exercises, one of the pair of interviewers took the lead in describing the problem and asking the follow-up questions, which allowed the other interviewer to closely observe and note the candidate's responses.

I was surprised to find that the interviewers did not give the applicant a chance to ask any questions - when I queried this it was clear that this was deliberate - with only half an hour available the interviewers were keen to maximise the time to use for assessing the candidate's aptitude and suitability. They did however check that all candidates were aware of the time and place of their second interview, and recommended that they talked through any queries they had with the student runners who were available to assist after the interview had concluded.

I was also surprised that the interview did not have a practical computing element. Again the interviewers were clear that they were not using the interview to assess an applicant's programming skills, in part because this differed so radically between applicants, but primarily because it was not something that was central to their suitability for the degree as taught at Oxford. Instead the focus was very much on testing and assessing a candidate's problem solving and reasoning ability and potential, which were viewed as much more fundamental than whether an applicant could demonstrate their ability at Pascal or C++.

After the candidate had left the room the interviewers discussed their performance and reached an agreed mark (out of ten) on a range of criteria using a report form in common use across the Department.

It was helpful for me to have had the chance to observe such a very different type of subject discipline from the humanities courses that interviewed last week. The approach used was clearly designed to assess the qualities required for the degree as it is taught at Oxford, with an emphasis on mathematical ability, application of theoretical principles and problem solving.

As a postscript, my immediate work colleagues at Essex bought me a separate gift. It was a battery operated watch. They had a very clear understanding of my potential and ability in mastering technology.