The powers that be, we hear, have set up a committee to "fix" entrance examinations for tertiary (read, college) education in science and engineering in India. It claims to be motivated by "the principle of inclusion for a collaborative excellence rather than exclusion through competitive excellence".
How one can avoid exclusion in an entrance examination is unclear. Isn't that, you know, the point of the whole thing? The Hatter put aside his initial misgivings, and read the proposal in detail. It does seem that a good amount of thought has gone into it, but not nearly enough.
It is all very well to reduce the number of redundant entrance examinations, but it's interesting to note that board exams (class 12) are being considered as replacements for entrance exams.
No amount of "statistical analysis" and "multi-parametric grading" can change a basic fact - when you have a college with such low accept ratios as IIT, you must have tests with extremely good discrimination between the 99th and the 99.1st percentile of ability. Yes, that good, even if that means making it impossibly tough for someone in the 90th percentile of ability. Any exam that can distinguish adequately between the 20th and the 30th percentile of ability will find it difficult to distinguish between 99th and 99.1st. Using board exams as an indicator is practically ruled out, since board exams are taken by everyone in the educational system, and cannot be targetted at the top few percentiles of ability. One way to avoid this issue is to increase the accept rates to much higher levels. That means setting up tens of IITs and a couple of hundred NITs, not just in name but in quality, overnight. That's not very likely, is it? Well, if that isn't going to happen, we have to live with needing high levels of discriminatory power in tests.
One can talk about two levels of examinations, but one can never use the same test that decides whether you pass or fail your qualifying grade to select candidates for IIT. In engineering parlance, the signal-to-noise-ratio at the high end of scores in board exams is insufficient. There's no way out other than having a tough examination, stressful as it is.
The one factor I really wish could be addressed is the effect of coaching shops on entrance exams. Even a couple of decades ago (when the ancient Mr. Hatter took his JEE), access to a good coaching center was considered important, and so was being in a "good" school in a metro. It was estimated that being a "small-town" kid with only access to "postal coaching" lost him a couple of hundred ranks at the minimum. These days, things have gotten even worse. Changing the examination pattern doesn't help this problem in the least.
The ideal selection test would be a tough exam that no-one prepares for. That, unfortunately is an impossibility.
How one can avoid exclusion in an entrance examination is unclear. Isn't that, you know, the point of the whole thing? The Hatter put aside his initial misgivings, and read the proposal in detail. It does seem that a good amount of thought has gone into it, but not nearly enough.
It is all very well to reduce the number of redundant entrance examinations, but it's interesting to note that board exams (class 12) are being considered as replacements for entrance exams.
No amount of "statistical analysis" and "multi-parametric grading" can change a basic fact - when you have a college with such low accept ratios as IIT, you must have tests with extremely good discrimination between the 99th and the 99.1st percentile of ability. Yes, that good, even if that means making it impossibly tough for someone in the 90th percentile of ability. Any exam that can distinguish adequately between the 20th and the 30th percentile of ability will find it difficult to distinguish between 99th and 99.1st. Using board exams as an indicator is practically ruled out, since board exams are taken by everyone in the educational system, and cannot be targetted at the top few percentiles of ability. One way to avoid this issue is to increase the accept rates to much higher levels. That means setting up tens of IITs and a couple of hundred NITs, not just in name but in quality, overnight. That's not very likely, is it? Well, if that isn't going to happen, we have to live with needing high levels of discriminatory power in tests.
One can talk about two levels of examinations, but one can never use the same test that decides whether you pass or fail your qualifying grade to select candidates for IIT. In engineering parlance, the signal-to-noise-ratio at the high end of scores in board exams is insufficient. There's no way out other than having a tough examination, stressful as it is.
The one factor I really wish could be addressed is the effect of coaching shops on entrance exams. Even a couple of decades ago (when the ancient Mr. Hatter took his JEE), access to a good coaching center was considered important, and so was being in a "good" school in a metro. It was estimated that being a "small-town" kid with only access to "postal coaching" lost him a couple of hundred ranks at the minimum. These days, things have gotten even worse. Changing the examination pattern doesn't help this problem in the least.
The ideal selection test would be a tough exam that no-one prepares for. That, unfortunately is an impossibility.
0 comments:
Post a Comment