The Zimbabwe School Examinations Council (zimsec) board is finally taking action, retiring the director and deputy director and sending other senior managers on leave, while there is a proper audit, but it took an earthquake in the form of a serious breach of security over an examination and a High Court case before this burst of activity.
And while the board probes its managers, its own members and those who appoint the board need to think seriously about the board's own deficiencies. This is, after all, not the first time a paper leaked. The credibility of an examination system depends on two major criteria.
The first, where zimsec passes with high marks, is the setting of good papers for examinations and ensuring these are fairly and properly marked. There has been little complaint about zimsec's educational professionalism and here the council gets a good grade, at least a B and quite possibly an A.
The second criterion is to ensure that there is no cheating, that all candidates have an equal chance to show what they know. This requires good security of examination papers from drafting to final delivery, as well as being able to replace papers that are compromised or even order a rewrite.
zimsec does not do too badly on the willingness to replace or order rewrites, so it has retained credibility of its examinations. But the cost has been high both financially and in public esteem because it has failed, and failed badly, in retaining total security of papers. It is all very well to be willing to clean up the mess, but there should be no mess to clean up in the first place.
Good security requires systems that work even if there is someone dishonest in the system. And zimsec needs to recognise that some people are dishonest. There are about 8 million to 9 million adults in Zimbabwe and roughly 20 000 of these are right now in a prison cell, one in 450.
While a significant fraction are incarcerated for crimes of violence rather than dishonesty, anyone creating a system that involves over 1 000 people -- and an examination paper once you include question setting, printing, distribution and a large number of examination centres must have that sort of number handling it before the examinees get their copy -- must assume that at least one or two of that crowd are seriously dishonest. With final approval of a draft paper limited to a handful of people, the dishonesty will almost certainly arise in the chain between that tiny group and the children writing the exam.
Some years ago, zimsec was very excited about a proposal to use locked boxes, one for each paper for each examination centre. These would have a combination lock, be loaded at zimsec headquarters and could only be opened a few minutes before the start of the examination when the combination for that examination was sent out by SMS, e-mail or WhatsApp.
You could even imagine two locks, with two people at each examination centre needing their own SMS to unlock the box. All the locks for each paper could be set by one person with the number in a sealed envelope in the director's safe.
The system would still be better than what is in place now. But modern technology allows other options. It is not difficult to create PDF files protected by serious encryption. These could be prepared for download, or even put on disk or flash drive, and sent to every centre.
Passwords would be sent out say 30 minutes before the exam so the papers could be printed on site, with examinees already inside their exam room to avoid a final leak.
These are just examples of possibilities. Many of these would cost very little to implement. What is required is a serious intention by the zimsec board and its management to find and implement a good system, one that is a lot better than what is used now.
That might well require a suitable security expert on the board, and almost certainly require a competent and bloody-minded security manager with a semi-independent remit to ensure security is taken seriously.