Main Logo

Tuesday, 28 February 2012

Z score and the concerned parent

I refer to the piece by a "concerned parent" on ‘The Z-Score and Professorial Pronouncements’ published in the Sunday Island on 12th February 2012. He is surprised that I did not consult Natha Deviyo on the Z score and tells the readers that, unlike the other Professors mentioned by him, Prof. Carlo Fonseka and I "have tended to obfuscate the issue and create confusion in the minds of the reader". I respect Prof. Fonseka more than most of the other Professors in Sri Lanka, though I have differences in opinion with him when it comes to Epistemology, the so called ethnic problem and Marxism, and I do not think he tended to obfuscate the issue and create confusion in the minds of the reader. Neither did I and as far as I am concerned I only tried to point out number of factors that both Prof. Thatil and others, and the members of the panel of experts appointed to propose a scheme of standardization had not considered. It has to be pointed out the panel of experts consisted of western Statisticians, I suppose of the same calibre as Prof. Thatil.

One of the factors I mentioned was the difference in the populations that sat for the examination in any subject in the old syllabus and the new syllabus. In any normal year the students who sit for the first time and the repeaters sit the same paper and are "pooled" in determining the Z score in any subject. This year as well as in 2001, in each subject there were two groups, one sitting for the first time and the other consisting of repeaters. The scheme that Prof. Thatil had introduced looks after the degree in difficulty with respect to different subjects and not with respect to the differences in the "merits" of the group of repeaters and the others in any subject. I believe that it is in order to look after this problem that pooling has been suggested by the panel of experts. I have no problem with pooling in principle, but I found that the "formula" that they had adopted was not quite correct. Having said that I mentioned that the differences in difficulty also had to be looked after and suggested after pooling and scaling, the students should be regrouped or "de-pooled" and another standardization should be carried out along the lines introduced by Prof. Thatil. It is not a recommendation as such and I am of the opinion pooling is more important than Thatil standardization.

In the articles I have read many had expressed the idea that populations of the two groups who sat for the first time and those who repeated cannot be pooled without giving any reason whatsoever as to why they cannot be pooled. In a normal year these groups are "pooled" or considered together and one has to come out with a convincing reason as to why they cannot be pooled without merely stating that populations cannot be pooled. I tend to think that in 2001 all the experts did not either consider the differences due to "merit" in the groups of repeaters and first timers or have a solution to the problem.

I did not consult Natha Deviyo in this case for the reason that I know that Natha Deviyo would not answer such questions, among others. Natha Deviyo is more concerned with people dying (two people per day) of Rajarata Kidney Disease and already he had given instructions on the cause of the disease, as western medicine has no answer to the problem. On the other hand there are western statistical methods to deal with standardization, a concept arising in the west, and I know that people are interested in solving this particular problem within the method of standardization, without waiting for us to come out with our own solutions. With respect to solving these problems in the courts of law I was reiterating a statement by Feyerabend on so called expert opinion in Science, with which I agree. It may be the accepted way of solving the problems, meaning the way imposed on us by the westerners how to solve such problems, but according to the traditions of my culture I question the so called accepted methods whether in law statistics or science, especially since we would not be independent unless we are independent of the knowledge of the westerners dominating our thinking, if we are thinking at all.

Copyright Prof. Nalin De Silva