Using IT to Improve the Education System

I hope the title is not entirely misleading. The talks by Prof. Phatak, head of Kanwal Rekhi School of Information Technology at IIT, Mumbai, were inspiring and made me realise that neither have I been writing as much as I had planned, nor have I made any efforts to alter that with which I disagree.

I recall a childhood tale about Akbar and Birbal. Akbar finds Birbal searching for something. Akbar asks him where did he lose it. Birbal points to another location. Akbar is confused and asks why then is Birbal not searching at that place. Birbal says that the light is here. Fortunately, in childhood stories, Akbar always learns the lesson. In the world of grown-ups, if I may mutilate a lovely tale by Prof. Phatak, we spend extraordinary efforts to solve the problem of a dead horse by trying everything other than recognising that the horse is dead and needs to be replaced.

In examination systems, we are computerising the release of marks, making them available on the web. We may even create systems which will monitor the coding of papers which presumably ensures that no examiner knows whose papers (s)he is correcting, or which examiner is not correcting fast enough and may delay the declaration of results, etc. We are always willing to spend enormous sums of time, effort and money on introducing a variety of checks to ensure the 'fairness' of the execution of the process, and hardly any effort or money on whether the process itself is fair. In general, the quality of the question papers is questionable.

I, like many others, was shocked by the paper leakage incidents recently. Et tu, Goa! Upon cooler reflection, it seemed obvious that people are the same worldwide and, statistically, placed in the same situation, will behave similarly. If people are not living up to our moral expectations, perhaps the problem is deeper than the symptoms suggest.

Our systems are based on suspicion as was pointed out by Khushwant Singh in a television interview over 30 years ago and also highlighted by Prof. Phatak just the other day. There is, hence, a total lack of transparency. The secrecy we introduce in the examination system ensures that no one is accountable for either the results or even the process. Roman Polansky's “Fearless Vampire Killers” ends with the comment that little did the professor realise that he was carrying away the very evil he had come to destroy.

A Suggestion for 10th and 12th Classes

Although I am convinced that our examination system needs a BPR (Business Process Review), let us consider a method which could be used for 10th and 12th class exams. We could have a fairly large question bank suitable for the curriculum of each subject created by teachers, students, and even any interested party. The question bank could be categorised by topic and difficulty, again by volunteers with a few dedicated experts guiding and monitoring the activity. Each succeeding year, the question bank can only improve because aside from adding more questions, we can also filter out the questions which are either answered by over 90% or over 90% cannot answer them – either case is useless for a certification exam. Testing is a very serious business and should not be confused with the routine, clerical processes.

The question bank could even be in public domain. If it is large enough and a student can answer all the questions, he obviously deserves to get full marks.

The rules for the question paper can be defined and a question paper generated from the question bank by a computer program. There will be no unpleasant surprise for anyone. No worry about which teacher is setting the paper. Will the students of that teacher get an unfair advantage? The question paper could be encrypted with the public key of the printer so that only the printer would be able to read the paper with his private key and print it. If there is a leak, we will know who is accountable.

Can't We do the same for Engineering Courses?

This is where I run into serious fundamental issues. What is the purpose of the examination? Is it a certification process or a part of the educational activity? The 10th and 12th exams have become certification exams because these are used for entrance options for various higher educational opportunities.

The certification of engineering students is being handled by GATE and GMAT for post-graduate studies in India, and GRE for studying in the West.

The University examinations must be a part of the learning process. They must inform the student and the teacher what the student has or has not understood. Simultaneously, the University needs to certify that a student has completed his degree and the class of a degree is to qualify the extent to which a student has learnt. The reputation of the University is at stake. This reputation is important for all the stakeholders. A degree from a disreputable institution is worthless for a student.

Hence, examinations are a necessary evil. However, they must be transparent. Each student must be able to see his paper. (S)he must be able to know how many marks were lost in which question and why. Once the examiners know that their work will be visible to the examinees and their colleagues, they are bound to be more alert about the quality of their corrections. The experience with the open source software and academics in general is a constant reminder that no better checks have been found other than peer review.

How can technology help in this area? The examination papers can be scanned after correction and posted on the web site. They should be encrypted with the public password of each student so that only (s)he can see the result.

But this is only the latter half of the process. We need to improve the process of setting up of the question paper. One example could be that a couple of days before the question paper is due, the teachers of that subject from each college could meet along with an Umpire and thrash out a common paper. The paper can be printed on a fast laser printer and be ready for distribution in a couple of hours. If we cannot trust our teachers with keeping the contents of the question paper confidential, how can we trust them with educating our next generation!

The teachers should meet, possibly, over the Internet using chat and email discussions to exchange notes and ensure that the syllabus considered by all the teachers is the same. Since the number of colleges is small, a substantial degree of freedom can be given to the teachers to make reasonable amendments to the course contents dynamically based on their experiences. Since they will set the paper, there will not be a problem of 'out of syllabus'. The results of the exams have to be used by the University to learn and improve its curriculum and standards.

Finally, I would prefer if we re-read the title with IT to mean Information and Trust. It is only then that we can help make Goa University one of the leading learning institutions in the world. We can and should change and change at a pace that we become a reference beacon for other institutions in the world.