My earlier article refuting the claims of hacking and marks tampering by ICSE board received more clicks than any of my posts so far. A section on privacy was quoted in The Hindu. The comments section is interesting with few people attributing things which I never said in the first place. Also, there is another article recently, claiming to “expose CBSE Board”. The reckless use of the words “tampering”, “fraud”, “exposing” is condemnable and can be distressful to many parents and students. At the risk of being repetitive, let me say that there is nothing so far to claim that the marks are “tampered”.
- Some readers mentioned that the missing marks pattern is a very rare event and one person even computed the probability (only he knows how) of such event to occur. It is important to remember that marks are not equally likely events. There is no reason why we should assume that the chances of getting a zero mark and 60 marks will be same in an exam. Also, some marks can be unattainable. For instance, if all the questions are of two marks each and if there is no partial marking, then the probability of getting an odd mark is zero irrespective of the sample size.
- ICSE council says that the board follows the process of moderation.
“In keeping with the practice followed by examination conducting bodies, a process of standardisation is applied to the results, so as to take into account the variations in difficulty level of questions over the years (which may occur despite applying various norms and yardsticks), as well as the marginal variations in evaluation of answer scripts by hundreds of examiners (inter-examiner variability), for each subject.”
- The process of moderation is very common across boards and it is good practice for a couple of reasons :
- Inter-examiner-variability : Some examiners may be very liberal in their evaluation and some might have the temperament of a typical Times Now journalist. It is very important to ensure that no student is at an advantage/disadvantage because of the liberal/strict nature of the examiner.
- Difficulty level adjustment : Assume that historically around 2% of the students who appeared for a particular exam got above 90% marks; if in the current year, the highest mark itself is 70% (assumption), then the board will do a readjustment to bring 2% of the students above 90%. This can be done across various intervals, using an algorithm without compromising the relative positioning.
- These adjustments, which are also referred to as “standardization” or “moderation” are very common across boards and it is a good thing because students no longer have to get panicked if the examination paper is very difficult.
- It is possible that after such a moderation, a given set of marks are not assigned to anyone, it could be in the intrinsic design of the algorithm that assigning some marks may not be possible. But, as I said earlier, as long as relative positioning is not disturbed, this is not a scam.
- Regarding the probability theory and bell curve, Why should everything look like a bell curve? Even if normalization is done for a bell curve design, How do we know that if marks are normalized or “intervals of marks” are normalized. For eg : It is one thing to says 50 students should get each of the marks between 71-80 and another to say that 500 students should be in the range of 71-80 marks. All this analysis will make sense only if the boards disclose their moderation methodology.
- Without understanding these simple designs of Education boards, allegations of tampering and fraud is purely attention seeking exercise. An eminent columnist sent me a message which roughly reads —“These people are thinking that they are exposing CBSE, but what is actually happening is that CBSE is exposing their knowledge of statistics”. That pretty sums up their conspiracy theories.
I am waiting for an RTI response on the actual moderation policy followed. I am not sure if the board will disclose it, but if it does, then I will upload it on the blog.