Search This Blog

Saturday, June 9, 2012

JEE 2013: IITK Senate Resolves

 The expectation was building up for the last few days. The media was predicting that IIT Kanpur Senate might decide to have its own separate entrance examination. And at last, it did. Every single speaker in the house talked about how the proposal is not just in violation of the autonomy of Senate, but it is academically and methodologically unsound. How, no due diligence has been done by IIT Council or Ramasami Committee. How there are contradictions in the decision, causing confusion all over. How, the IIT Council proposal will increase coaching, and spoil school system. How it is going to increase the stress amongst the students.

Senators wondered how the HRD Minister could pronounce the decision to be unanimous when it violates the Senate opinions from pretty much all IITs. One senator read the Senate resolution from all other IITs, and showed how most resolutions were far away from the final proposal minuted by IIT Council.

It was one of the best attended meeting in my memory, and considering that it happened when half the faculty is actually out of town, it really showed how the faculty is passionate about preserving the quality of education and the autonomy of the Institute. It really was an amazing meeting, where senators had come prepared, read all the background papers, remained very focus on the issue, and spoke eloquently about the disaster that IIT Council is trying to create. But at the same time, the senators showed exemplary restraint and did not make any personal remarks about anyone in IIT Council or Ministry.

Here are the resolutions passed in the Special Senate Meeting of Friday 8/6/2012.

---------------------------------------------

The Senate of IIT Kanpur resolves that:

1.
(a) The recent IIT Council proposal on admissions to IITs is academically and methodologically unsound and is in violation of the Institutes of Technology Act (1961) and IIT Kanpur Ordinances (Ordinance 3.2 (Admissions)).

(b) In view of the Senate resolution adopted in its 2011-12/9th meeting held on April 10, 2012, invoking Ordinance 3.2 the Senate resolves that IIT Kanpur will conduct the entrance examination for admissions to its undergraduate programmes in 2013.

(c) Given the present exigency, the Senate authorizes the Chairman Senate to constitute a Committee with the help of DOAA for conducting JEE 2013 by IIT Kanpur.

(d) The Undergraduate Admissions Committee will organize the entrance examination and counselling, and deal with all other matters pertaining to the undergraduate admissions in 2013.

(e) To the extent possible, the committee will coordinate with other IITs to conduct the entrance examination jointly.

2. In order that the views of the IIT Kanpur Senate on the issue of undergraduate admissions are represented in the public domain adequately and accurately, the Senate authorizes the following senators as its spokespersons till further order: Profs. Y. N .Mohapatra, Harish Karnick, Somenath Biswas, H. C. Verma, A. K. Chaturvedi, C. S. Upadhyay and Deepak Gupta.

3. The Senate reiterates the stand it took as represented by its Chairman in the May 12, 2012 meeting of the IIT Council. Consequently, the Senate resolves to record its forceful dissent of the Council resolution related to JEE. This dissent may be recorded by a letter to the Chairman, IIT Council and Chairman, BOG from the Chairman Senate.

-----------------------------------------------

What next:

 In pursuant to this Senate resolution, Director has already formed the Under-graduate admissions committee, under the Chairmanship of Prof. Neeraj Mishra, Department of Mathematics and Statistics.

The first and immediate task of this committee will be to coordinate with other willing IITs on the issue of conduct of the admission test in 2013.

One of the media person asked me whether it would not be very stressful for the 12th class students if each IIT were to take a similar decision. My answer was no. There is likely yo be two sets of IITs, one which go for the new JEE on the basis of the model proposed by IIT Council, and the other, who will join hands to conduct their own admission, as we have been doing for the last 50 years.

A large collection of documents regarding this JEE saga is available at this link.

177 comments:

kasara said...

Hats off! I never expected this to happen in India.- I mean, debating a proposal put forward by the government and trying to engage them in a healthy discussion. IIT Kanpur , I hope, becomes an example for other educational institutions in being assertive.

Siddharth Jain said...

Great work Sir, Hats off to you and your colleagues at IIT K. You showed the guts.

Unknown said...

I hope all iits join hands with IIT-K
as I am directly being affected by the decision of the IIT council.

Unknown said...

Recently wasn't there a confusion about the autonomy of senates of older IITs Vs the IIT Council ? I mean how can IITK conduct its own admission examination legally? I am not much aware of the IIT Admission act.

Dheeraj Sanghi said...

@Sumit, IITK Senate believes that it has the necessary autonomy under the Act to conduct its own entrance examination. At IIT Kanpur, we will keep doing things as per our understanding of the law.

Prashob's blog said...

Dear Prof.,

I sincerely appreciate IIT K for standing up against this bull dozing of so called "reforms" by the government.

However, going by the way people make policies in this country, the very Institute of Technology Act might get amended by the short sighted politicians to shove this reform down everyone's throat.

Hope this battle of David Vs Goliath makes our institutions and the country better. I (its immaterial) fully support this resolution.

My best wishes.

Devesh Tiwari said...

Is there a danger in future that HRD minister puts some clause in IIT Admission Act that denies IITK to conduct its own entrance exam?

Prashant said...

Great - Congratulations ! Academics should not be shoved around by ministers and political stooges who represent them in the form of directors.

Anonymous said...

Thank you Sir! Great to know that their are people who would stand up to fight for autonomy of IIT. Looks to me Mr. Sibbal think most serious problem in Indian education is IIT-JEE and IIT directors "genuinely"(As suggested by Dr. Barua) agree with him.
As I understand from point of view of those who are high and mighty:
1> It just happened that all IIT director and Minister are of same view. How does it mater if other differs. We are powerful so it is our way
2> Central Motive of JEE: JEE is responsible for degradation of education system in India. The moment you make 12th marks a player in JEE admission all of a sudden state boards will start to reform itself.
3> As stated in Prof. Barua's article in Indian express: An important goal of this reform is to make coaching irrelevant. However, he agrees that if one take near term view: coaching institute will do killing with current change.
4> Somehow the current changes in JEE will do something which will force coaching institutes to disappear in long term.
5> I feel current system has flaw. So, only way to cure it is to destroy it.
6> People who are defending quality of CBSE over other boards are doing so because their ward study there.
on his last Gyana: Another anecdotal fact people who make maximum noise against JEE are those whose ward fail to make into IIT.

Manish Verma said...

This should send a strong message against top-down approach. There are some points as per IIT Act (http://www.iitm.ac.in/downloads/RTI/Act.pdf) though that are worth considering,

(i) As per 13.1, the board has the power to review the acts of the Senate.
(ii) As per 13.2.e, the board of the institute can cancel or modify ordinances.

Note: As per 11.e, the board will have two professors of the Institute, to be nominated by the Senate besides other people as mentioned in 11.a, 11.b, 11.c & 11.d.
(iii) As per 9.3, the President of India may take such action
and issue such directions as he* considers necessary in respect of
any of the matters dealt with in the report (refer 9.2) and the Institute shall
be bound to comply with such directions.
(iv) As per 33.2.a & 33.2.b, the Council itself has some powers especially with regards to laying down policy in the current context.

If various parties are convinced that they disagree and the debate/battle will continue, would it not be better just for 2013 to screen students based on x times calculation (x > 5) to minimise error due to error propagation and select the students on the basis of JEE-Advance alone.

* It is, "he" alone that is mentioned incidentally and not he/she.

Shishir said...

Wow!!!
Of course, the some battle is ahead.But,if one has to go down , it's noble to go down fighting, more so for a just cause.

What is life if we can't even stand up for what we believe in .

I'm really happy to be part of IIT/Kanpur.

Shishir said...

Dheeraj, I'm happy to hear that IIT/K senate believes it has the necessary autonomy and it'll keep doing things as per this understanding.
But, given my understanding of the dynamics of MHRD (or of any ministry) the matter wont end here ( I need not elaborate!).
Therefore ,it is necessary to continue to emphasize in various quarters and forums (lobbying , I keep saying, is not a bad word)that Academic independence is essential for blossoming of young minds. We are lucky to have institutes like IIT which have the critical threshold level of academic excellence necessary for the brilliant young minds to share,interact and grow upon it to compete with world's best. Young minds also need freedom to really contribute originally. And financial control should not necessary mean academic control.

IIT system's independence and academic freedom is in nation's interest. This needs to be put across to various stakeholders, including MHRD and the government. This will require some effort of non-academic kind, but , to my mind , it's absolutely necessary as no one else is going to bat for you.

Unknown said...

@all : I am not able to understand the stand of IITK director prof dhande.... He gave presentation in favor of new scheme in the meeting with states as media report suggests.... And he is also the chairman of senate ??

Dheeraj Sanghi said...

@Manish, We are aware of the fact that Board has powers to review all acts of Senate. But two things. One, we are only saying that the process of admission which has already been approved by board decades ago will continue for 2013. So board does not come into the picture. Two, we have talked to some of the board members, and we are convinced that they also love this country as much as we do, and would like to maintain the excellence of IIT Kanpur.

Dheeraj Sanghi said...

@Shishir, Thanks a lot. What we have been miffed about is that even small things that we ask for are not accommodated by IIT Council. We have the same goals as they say. We have also clearly said that we are willing to try their model. We are only asking them to go slow and collect data on the way, and make incremental changes over the next 3-4 years, assuming the data in these years show that their assumptions are right.

If even such a simple request by a majority of IIT Senates can be rejected by IIT Council, then there is not an iota of autonomy left.

And this is not the only agenda on the table. We are already hearing attempts to have same curriculum being forced on all IITs. (I am told that some meetings have already taken place without the Minister, but so far Directors have opposed this. But tomorrow Minister will be present in one of those meetings. Will Directors still say no to common curriculum.) So, if we do not fight now, immediately, even curriculum will be decided by the Ministry.

Satpathy Amiya Kumar said...

IIT standards must never be compromised by petty politics of Congress and Mr. Kapil Sibal. Indian IITs are seen with great esteem all over the world. The petty politics of Congress and Mr. Kapil Sibal can never be allowed by all means.

My salute to IITK.

saahil said...

IIT-k has taken extraordinary step in the right direction.Kapil sibal had himself said '' IIT's have some concern's which I don't understand ''. Instead of addressing those concerns, he wanted to become a posterboy for change and went ahead with his destuctive plan.

also saw your show at NDTV and how proff BARUA turned down the serious questions posed by a parent from dubai..

gautam barua said...

The IIT Council is asked to be reasonable and to accept "simple" requests. But what is this simple request? Do nothing for now and let us look at "data". What data? As was pointed out in the TV programme, data is available, if you look for it and ask for it. It is not fair to ask the IIT Council to supply you data to convince you. Isn't this being arrogant? Who is hiding what data? And what will you do with data? Show that CBSE and ICSE kids do much better than other Boards in IIT JEE and other exams? This is already known. We do not need new data to prove this. But the question is why are they doing better? This debate has already taken place, and I think it has ended inconclusively (in Dheeraj's blog).
On the other hand, the IIT Council, after getting the inputs from 7 IITs, made major concessions and proposed a new scheme. If you look at IITK's Senate feedback and the new format, the following seems not to have been accepted: 1) 2013 instead of 2014, 2) the second IIT only test being given to everyone instead to only a shortlisted set. Now, take a look at what the other IIT Senates opined. If you look at all of them together, you will find that the compromise formula has met "most" of the suggestions. Isn't that what a compromise is all about? It cannot clearly mean the meeting the demands of one of the 7 units. In fact, the compromise is identical to the option 1 of IIT Madras. So, isn't there a case to state that the IITK Senate is objecting now without clarifying what is the basis of their objection? 2013? Let us have a debate and I can (as I have elsewhere) tell you how the compromise formula can be easily implemented in 2013, while the earlier proposal was difficult to do so (and IITG had suggested 2014 for that proposal). But it seems it is not about "implementability", it is about "data", it is about "autonomy", and it is about "board marks" (after accepting the proposed change from 2014 in their earlier feedback to the Council).
Was there a loss of memory? Or was it a decision in the heat of the moment, based on having felt "cheated" ?
I sincerely hope there will be a re-consideration.
Please let us not create permanent harm to the IIT system.
And, please no Director bashing just for the sake of it!

Dheeraj Sanghi said...

@Gautam, The most important pillar of your argument for 12th class that it is possible to compare boards has been proven wrong. For months, you kept defending it. Now, you can no longer defend it. So you are saying it that you knew this all along.

Since you are the most vocal and indeed the only member of IIT Council who is willing to say things publicly, if you are saying now that you knew something all along and you chose to ignore it, what trust should we have in the proposal.

There is a body of research that I have come across now which shows that school education in TN has suffered because of 12th class marks being the sole basis of admission.

And, if you had data, and you have been hiding it, then you are responsible for the strong reactions that we are having. Now, you are essentially admitting that you have been unfair. I wish you had admitted this 3-4 months ago.

You think there is absolutely no issues with others being forced to give the second exam. We feel that that is a major issue. If people are forced to give an exam for admission to some institutes by the government, then those institutes MUST have some say in the conduct of the exam. That is only being fair.

gautam barua said...

Dheeraj,
I have not admitted anything of the sort and I have not hidden any data. I wonder how you have drawn the conclusion that the "most important pillar of comparing the Board results" have been proven wrong.
As far as the Tamil Nadu Board issue is concerned, their problems, in my opinion, is that they did not use rankings and instead went in for absolute marks. It has become very difficult for CBSE students to get admission in Tamil Nadu. Further, they have used ONLY Board marks. All proposals made so far by the IIT Council have not suggested this.
I am not interested in winning any arguments, and I concede that you have an opinion, but I disagree with it. But the issue now for me is to convince IITK faculty to reconsider their decision and save the unity of the IIT system, nay, even the IIT system.

Ankur Kulkarni said...

"Show that CBSE and ICSE kids do much better than other Boards in IIT JEE and other exams? This is already known. We do not need new data to prove this."

Prof Barua, this is very disturbing. What was the basis on which you were equating percentiles across boards earlier?

Shishir said...

@Gautam, I heard you at TV among other panelists including Dheeraj. I believe that one of the main reasons offered for using or giving weight to board marks (or percentile )for IIT admission is to improve board's quality by way of forcing the students to pay more attention to their school education. [This is what I have understood in all these days ]
But,there was a time,and I am quite sure you remember your time, when admissions (barring IIT and Roorkee) were based solely on boards marks . This included RECs also. This was stress free too. (I just had to mail my Intermediate mark sheet to various colleges and I got a congratulatory letter confirming my admission.That simple.)
This practice had to be abandoned later as correlation between a student's boards' marks and his performance in college came down.
Now , reverting back to the old system , what is being sought to achieve? Could you please elaborate?
I may add that in the name of making Board Exam stress free , most of them introduced MCQs or short questions which increased pass percentage phenomenally. In UP, it went up from a 17-20 % to 85-90%.Earlier, it was almost given that UP board topper will have cleared JEE with high AIR. Now , it is almost the opposite.
Don't you think that given the shape most of our state boards' are in (they have not gotten any better), reintroducing weights to boards marks will fuel the very problems which necessitated entrance exams in the first place?

Ankur Kulkarni said...

Prof Sanghi: Congratulations to IITK for taking such a firm view. I don't agree with you objections, but I admire the assertiveness that your senate has shown. It is also a sign that IITK is not fearful of any backlash from the ministry and is secure and sure of its standing.

Anonymous said...

@Gautam

Respected Dr. Barua,

I request you to please answer my question.

Suppose the new system proposed by the MHRD and the council, publicised as "One Nation, One Test" had been in place in the early 70s when you were preparing the for the JEE and if after studying hard for months, on the day of examination you would have fallen ill and wont have been able to perform well, what would your life be then? Would you consider such a system to be just, which clubs the exam for the best institutes of the country into one day without keeping in mind the ground realities?

This would definitely be a minor issue in your eyes but not for the millions of students and their families.

Piyush P Kurur said...

Dear Professor Barua,

I am surprised that you are miffed with the fact that many are demanding data to support the claims you are making namely

(1) That the new formula will reduce coaching (there is data to show this will not happen. Check the ads of coaching institutes)

(2) That the stress would reduce.

(3) That school education will magically improve.

Let me quote Carl Sagan:

Extraordinary claims required extraordinary evidence.

You can shift the burden of proof to the skeptic. That is not science. Atleast as I know it. Sorry


Piyush P Kurur

Unknown said...

Sir,
i hope to bring to light a major flaw in this current test format which may have been overlooked.i am currently pursuing engineering in iit-bhu(going to be 2nd year).Last year a girl topped the boards in my school and she went for medical.According to the present test format she is having a fair shot at this JEE although she has a different aspiration.She is opting for one of the 50000 supposed seats which could have been a seat of an engineering student.So that seat is WASTED.The government doesnt see this.After all not all students pursue engineering.

Saswata said...
This comment has been removed by the author.
Anonymous said...

"The most important pillar of your argument for 12th class that it is possible to compare boards has been proven wrong."

When? By whom? Via which publication? And, most importantly: How? With respect to what considerations? what evidence?

Ajit
[E&OE]

Anonymous said...

@Ankur:

You asked Prof. Barua the basis on which he has been equating percentiles across the boards.

I don't know about him, but I would say that it is: the identity of form of the distribution functions for the different boards, something which is ensured (at least in the statistical sense!) by the law of large numbers.

When CBSE boards folks do better on JEE than, say, Pune Board folks, what it means is that the "composite" product of {student + syllabus + schooling + higher social stratum correlates including possibly a second-hander's motivation to have to do well at the JEE, and money-related advantages such as affordability of high-end coaching etc. is better for the CBSE board student than the Pune board student, when JEE is used as the yard-stick. Sorry for the awkward expression, but I do mean to say that everything in the curly braces {} goes into that composite product which is tested by the JEE (to whatever extent---good or bad), not just the student or his basic (even "innate") intellectual abilities.

BTW, a lot of debate would get simplified if everyone used the right concepts: probability distribution function---its form.

Another relevant concept: repeatability. There is ample anecdotal data to believe (against which Dheeraj won't argue) that more preparation, including repeated attempts, improve JEE scores, esp. in the top range. (People take pride in getting into CS branch right in the first attempt, in deference precisely to this fact.) Now, my point: Any test that allows as much improvement in scores at the very top end as JEE does, and any test that allows as much ranks improvement with coaching in the top and middle of the topmost range as JEE does, simply fails to be a good measure of the basic (even "innate") intellectual abilities of students, at that range (viz. near the top).

In essence, Prof. Sanghi argues for preserving the bias of the IIT UG admission process towards social endowments more than the basic individual abilities. That's what he, in effect, ends up doing, though he probably does not intend doing so (and, probably, does not also realize that that's what he ends up doing). In contrast, Prof. Barua ends up better supporting the basic individual abilities, perhaps unwittingly also in his case. (Sorry if I have insulted both of them; I didn't mean to. I was just trying to elucidate what their positions, effectively seem to be like.)

Ajit
[E&OE]

gautam barua said...

One of my comments have not been posted!
I am responding to the issue of everyone being "forced" to take the advanced test even though they are not interested in admission to an IIT. I was unable to put forth this point in yesterday's TV programme.
I had proposed in the May 28 Council meeting that NITs could use only the Board marks and Mains result for their admission. But the NIT Directors wanted NIT admission too to be based on the results of the advanced test. Clearly, it is an issue of branding for them. So complaints on this front may please be addressed to NIT Directors.
In the posting that was not put up: I refute that my earlier post implied that Board marks cannot be compared based on percentiles. I do not concede any such thing.
I appeal to the IITK faculty to reconsider their decision.
Gautam Barua IITG DIrector

Prashant said...

" Who is hiding what data? And what will you do with data? Show that CBSE and ICSE kids do much better than other Boards in IIT JEE and other exams? This is already known. We do not need new data to prove this."

This statement, coming from an IIT Director should ring an alarm bell. What kind of approximate calculations are you using to determine this new system ? Isn't that exactly why you should give data to those who are asking for it ? We all agree that schooling is important but if you factor it in with 50% weightage even for percentiles, it is unacceptable if you were going to throw it in, without knowing how much it would swing the results. Otherwise, leave it at 20-25% weightage, or perhaps use it as a sufficiently wide-band filter, picking up the top 20-25 percentile. So that schooling gets importance, but not enough to throw in an unknown swing.

But then perhaps this is not so much of a hasty plan, but by design, supposed to be some kind of indirect state wise quota in the IIT System which might fetch brownie points for someone in an election year.

While I don't think of even the current JEE system as all that awesome a system I think IIT Kanpur has done something extremely creditable by sending a message to the directors that they are not going to be shoved around by MHRD implants. It looks like other IITs are to follow. I was reading a video of the IIT-KGP director supporting the HRD move and saying that KGP supports it. Below that I saw comments from the faculty rubbishing his claims.

Prashant said...

From IIT Kharagpur's main Facebook page - looks like a storm is brewing there as well.

The IIT Kharagpur administration allegedly created a false impression that its faculty supports a Union government proposal for a new joint entrance exam for the IITs that has triggered fierce internal turmoil inside the institutions.

Several IIT senate members have said the HRD Ministry is trying to impose the new exam pattern on the IITs with little debate, lack of consultations with faculty — through a process facilitated by IIT directors who have ignored faculty views.

Manjeet Dahiya said...

Dear Prof. Sanghi, Heats off to you and IITK Senate!!
Please continue to fight, all the other IITs are going to join forces with IITK and continue to have the JEE like we had 50 years.

@Prof. Barua I saw you on TV and personally found you being illogical. You answers were vague and irrational. E.g. the one of having two exams the same day. You compared it with the B.Arch. exam. The Arch. exam is taken by few and just to prove your meaningless point you compared it with JEE-Advance. I request you to join hands with the other IITS.

gautam barua said...

1. Sorry, no post is missing. It came up late.
2.@manjeet: Please see my comment of 12:22 pm re: the second exam.
3. @ajitjadhav - unwittingly or otherwise, thanks for elaborating what I meant when I mentioned that "Show that CBSE and ICSE kids do much better than other Boards in IIT JEE and other exams" !
4. @prashant - the issue of "state wise quota" has been already been discussed in many other places. Please believe me when I tell you there is no design of the kind you mention. In fact, let me repeat the basic arugment: the ISI Kolkata report implied that the only way of ensuring correlation among Board marks which will not vary from year to year, is the one based on percentile ranking. Any process based on normalisation of actual marks will invite feedback from Baords by way of inflation of marks, changes in distribution, etc. You do not need data to assert that ranks will not be susceptible to such feedback. If you agree with what Ajitjadhav has said, how can past data be used to prove anything? If anyone can suggest a sounder way of comparing Board marks, I am sure it will be considered very seriously indeed.
5. Please also remember, there is no best solution. The new proposal can be criticized, as can the current system. Surely no can assert that the new scheme will solve all problems in schools, or completely stop coaching, or be completely fair, and will be guaranteed to be better than what is there. We must learn as we go along, and be ready to make changes if we find that assumptions have gone wrong. We should, however, be ready for change.

Gautam Barua

Shishir said...

@Gautam, It appears that you have missed the essence of the counterpoint.
Given the present state the state education boards are in, the percentile and absolute marks don't make much of a difference. As I said earlier, state toppers ( some data can collected easily) don't easily crack JEE or AIEEE and their performance later in whichever college they is not something to write home about.
This clearly shows that 'percentile' thing will also throw up wrong kind of weightage and will discriminate against good students.
The solution lies in correcting the state boards first.First and foremost, make them fair. The toppers need to be real good. That's how the system should work. IIT gold medalist will be accepted anywhere in the world , by any yardstick. This is the real value of IIT system. The moment IIT throws up toppers or 9 pointers of iffy quality, the system will no longer be worth its name.
It surprises me to notice that education administrators like you fail to see such a basic thing and failed to give a proper feedback to the ministry and the minister.

PULKIT said...

Respected Sir,
I have a doubt. I am quoting few lines from Institutes of Technology (Amendment) Bill 2012,
"(m) until the first Statutes and the Ordinances in relation to the Indian Institute of Technology (Banaras Hindu University), Varanasi are made under this Act, the Statutes and Ordinances as are applicable to the Indian Institute of Technology, Kanpur immediately before the commencement of the Institutes of Technology (Amendment) Act, 2012, shall apply to the Indian Institute of Technology (Banaras Hindu University), Varanasi".
Now legally Ordinance 3.2 of IITK, which talks of a separate test, must be applicable to IIT BHU as soon as the Act is published in the Gazette. So can it be inferred that IIT BHU would automatically come under the new test by IIT Kanpur?

Dheeraj Sanghi said...

@Pulkit, it is the interpretation of each Senate. We believe that the new JEE is not what is meant by our ordinances. IIT BHU may decide that the new JEE is indeed the JEE being talked about in the Ordinances. And we have made a resolution to make it clear to all stake holders that the old JEE is what the ordinance should be interpreted as.

Unknown said...

@prof Barua
I passed my 12th this year with 84%.I am not satisfied with my rank in IIT-JEE.I already had a plan to drop a year and try once more.And one fine morning you people tell me that there is no IITJEE,no AIEEE. A single test. And the most important part 12th marks going to be added !
Now tell me with 84% what percentile would a student get?And you tell so freely that i can write my boards again as if it is a walk in the park.And can you please explain the reasons why Mr HRD does not want to implement his scheme in 2014?Does not make any sense.

Ankur Kulkarni said...

@ajitjadhav, Please take some time to read my probabilistic analysis here and show how it may be modified to mathematically make your claim.
http://www.blogger.com/comment.g?blogID=4273139713770751485&postID=5087225485998447665

I was initially under the impression that the central limit theorem will provide parity across boards for the same percentile. After doing the math I realised that this is not obvious and at the very least, some scaling is required.

aaa said...

gautam baruaji,
can I request u to give logical reasons for not acceding to the request for change with effect from 2014. Are u and Ministry of HRD not aware that a student starts preparing for IIT JEE or AIEEE latest in 10th standard. Some do start much earlier. Such is the level of competition. And now u are supporting a decision of Ministry of HRD wherein a new unproven and un-analysed system is being implemented in the middle of Standard 12 when students have all along studied for a particular pattern. Is it not unfair in your views? And moreover, students who will be taking their second chance in 2013 have been asked to take board exams again if they so desire> Well, in your view, it may just be a matter of filling a form of CBSE and just appearing in the exam. How naive, unrealistic and unthoughtful and emotionless decision. Instead of all these changes, would it not be better (assuming percentile issue and other issue are found implementable and fair) to have these changes implemented with effect from 2014. Why it is not being understood that a good and sound change will be welcome by all. So why it can not be done that the new system is implemented by 2014 and by say December 2012, all the issues are discussed and consensus reached thereafter in a transparent and faithful manner for their implementation. What is the big hurry, please tell. Are u and the ministry at all concerned about improving the school education system as a whole or u think that by fiddling with the exam system, all foils of ministry of HRD about pathetic state of schooling system will be forgotten. It is very scary to listen to the logic being forth by such intelligentsia just to support a worthless idea, for some unknown reasons. And I am not talking about many issues about board exam results, percentile, psychological effect on students and parents alike, only one chance in a year, exact pattern of JEE advance etc. And finally, this magic figure of selecting only 50,000, just how did it come about. I hope u and the ministry realises that as at now, only 5-6 lac students appear for IIT. With so called new system, 12-13 lac students will be appearing with vast difference in their curriculum, evaluation, cheating issues, inflation issues and only top 50,000 students will be considered for IIT JEE advance. I am sure, 80% of the deserving students would not even make it to top 50,000, courtesy this new system. What an idea, sirji? I request u to respond please

Piyush said...

ajitjadhav said:

....the identity of form of the distribution functions for the different boards, something which is ensured (at least in the statistical sense!) by the law of large numbers.



I am sorry, but it is not. As Ankur Kulkarni pointed out above, the Central Limit Theorem has two caveats: the limiting distribution depends upon the sample variance, and it depends upon the numbers being large. At least as far as I can see, no one is putting forth any data to show that the variances across the different boards are comparable. Given that, fixing the same percentile (say 90th) for all the boards would be simply ridiculous.

ammu said...

Respected,
I am proud to be an IITK alumni, and very fortunate for choosing IITK during counselling and not IIT Guwahati. The assertiveness displayed by the IITK fraternity for guarding its autonomy and rejecting the ill conceived, politically motivated decision of few for the interest of students of this nation will serve as inspiration for other intellectual bodies. My salute the fearless academicians of this great institute.

Manish Verma said...

Regarding,(b) In view of the Senate resolution adopted in its 2011-12/9th meeting held on April 10, 2012, invoking Ordinance 3.2 the Senate resolves that IIT Kanpur will conduct the entrance examination for admissions to its undergraduate programmes in 2013, as per http://www.iitk.ac.in/infocell/iitk/newhtml/ordinances_03.pdf, The Admission of Indian Nationals to the B. Tech., B. Tech.-M. Tech. (Dual Degree)and M.Sc. (Integrated) Programmes shall be made once a year on the basis of the Joint
Entrance Examination (JEE) conducted jointly by all the IITs. It says, conducted jointly by all the IITs, it is not clear how can the Senate of a particular IIT decide to conduct JEE for its own institute.

Manindra said...

@Gautam: you are making three points. First, that you and Dheeraj have a difference of opinion, and that two of you should agree to disagree. Second, that IITK faculty should agree with your opinion to save IIT system. Third, that even if we do not know where the proposed changes will lead, we should give it a try.

Let me suggest a better alternative: YOU should agree with IITK faculty. Why is it better? Firstly, our opinion is backed up with better evidence than yours. Secondly, it will still save the IIT system. Thirdly, the current JEE exam, for all its warts, still delivers on a fundamental objective: it is very fair. There is a good amount of evidence that in the proposed system this objective will not be met. And change for the worst is never a good idea.

Anonymous said...

Dr. Barua, what kind of academician are you? Everybody knows why directors are supporting IIT council decision. IIT-KGP is the latest example where director over-ruled its senate's decision. The whole nation looks upto you. Don't be puppet in hands of Govt. and for God sake stop playing with future of lakhs of aspirants only to satisfy your ego.

Anonymous said...

Dr. Barua, as head of an institute of national importance, don't you think you should be more careful before making a comment in public. Unfortunately your view point is being endorsed as viewpoint of entire IIT-Guwahati. But really? Let us know what your senate says, not what you feel.

IITKblogger101 said...

Two contrasting faces of the iitg director:

How is the coaching culture affecting students?

It makes its impact on students after their selection. On entering the IITs after undergoing excessive coaching, the students are almost burnt-out and mentally fatigued. Then the IITs for them become a place to relax. Coaching is the primary reason that affects the performance of students. A mindset has been created that if they (students) do not opt for coaching, they may not have the chance to get admission to the IITs. That is why there is such a big business in coaching.

IIT has a brand value. Parents and students want to enter IIT without thinking what they will do. So after they graduate from the IITs, many of them don't go for engineering jobs; they rather go for the finance sector, management or do jobs which have no connection with engineering. We have seen that about half of the students from IITs are really not interested in engineering.



and

http://www.impacteducationcentre.com

Sriram said...

I don't understand how you can equate scores across different board exams based on percentiles. The law of large numbers does not work here because the populations that take the respective board exams differ in the first place due to a variety of reasons. Dr. Sanghi is correct that the entire Tamil Nadu system has been progressively trashed by awarding absurdly high "marks" and consequently reducing the distinctions among the more able students to be just statistical noise.

Saswata said...
This comment has been removed by the author.
Unknown said...

Looks like Prof Barua is never coming back to this page...

Rohit said...

Comments left by IIT-G Prof( Saswata) here clearly shows that Prof Barua was willing to change the meeting minutes when 90% of IIT-G senate were against his views. Amazing, simply amazing. I will love to hear his arguments for doing so. Probably he thought that wisdom of 90% of faculty at IIT-G was meaningless so why not change the meeting minutes.

While browsing, I saw Mr.Barua pictures of attending a coaching institute function. He claims that it was best way to project IIT in north east. He could not come up with any better way to promote IIT in North East.

See the blog entry and his comments,

http://iitproblems.blogspot.com/2010/05/dual-face-of-director-gautam-barua_22.html?showComment=1339379600361#c6805163064883643120

Devesh Tiwari said...

@Dr Barua:

Everyone is asking you for data. I am not going to ask you that. You will say data contains noise and noise removal is non-trivial etc. So let's only will talk in terms qualitative arguments. I have asked this question before but it was in another post so you may not have noticed it. Here I go again.

Since my arguments do not include data, others may find it vague and unscientific -- apologies in advance.

How is your proposed system going to tackle State Board level inefficiencies?

As I have pointed out earlier, that if you only change the set of examiners in State Board exam, the order of students in top 10%-ile of State Board will change very significantly. That is because State Board examiners (e.g. UP Board) are not concerned about fair and consistent grading, they only care about throughput, i.e. number of copies graded per day (directly correlated with their reward/payment).

Your hope, as I understand, is that this will improve with time as State Boards will realize that IIT-JEE (significantly) depends on their grading quality. Hence, state government will put more pressure and have better scrutiny system in-place.

I hope the same as you do because it will eventually improve the State Board system as well. But let me articulate why it won't work.
The following argument is mostly qualitative to be fair.

The reason is the fundamental mismatch between the purpose of State Board and your hope . The fundamental purpose of a State Board (e.g. UP Board) is to give 10+2 certificate to lakh and lakhs of students in time. Getting all copies graded in time, their focus is never on consistency and fairness. They believe their fairness is good enough, their only thrust is getting things on time.

My dad worked in Allahabad office of UP State Board, and during all his service years, everyone just talked about getting things done anyhow, and that's it. That's the only thing they can achieve, it's not that they don't want to achieve 100% fairness and consistency. They say they can not achieve 100% fairness and consistency because of huge number of students appearing in the exam. Most of these students want just a degree certificate, only a small fraction of students give a serious shot for JEE.

You can see that State Boards have a bigger cause to serve and are not going to make their system more complex and time-consuming just for these small section of students going to appear in JEE. Moreover, what incentive does an examiner have to be 100% fair and consistent when he is being paid 100Rs. per copy in Allahabad's heat ?

To summarize, I showed that top 10%-ile of a Sate Board is so unreliable and inconsistent that they are not going to serve your purpose.

To reiterate, you purpose is: while deciding whom to admit in IIT give higher weight-age to students who did better in State Board exams. Here we just showed that who did better in State Board exam is itself not reliable.

Thefore, this system is not fair to State Board students either, forget about this system not being fair to CBSE students.


A State minister can get his kid in top 20 merit list of a State Board, and try to push his chances for IIT.
Worse, each minister has five such relatives who are appearing for 10+2 exam. Remember, quality control is in State Govt's hand.

A corrupt state government may trick the JEE exam now, which was not possible before. Whom should we thank for this?

Thanks,
Devesh

Unknown said...

@Gautam: I have had the privilege of encountering you in a meeting at IITK in which you were a representative of the MHRD team comprising of three honourable directors and Mr. Anandkrishnan this year in April. I can further characterise the meeting as the one in which a few members of the MHRD team lost their cool after the rationales behind the new "scheme" were asked for (I know you are good at denying facts, mincing words or even worse the need for facts, but this happens to be an objective reality and you'll only invite ridicule if you do so!).

I would only request you to clarify:

a) your insincerity with regard to preparation of the minutes of the meeting as reported by Prof. Saswata.
b) your dual standards with regard to coaching, as reported by IITKblogger101 and Rohit.

@Prof. Sanghi - standing ovation for being firm in this issue which has now gotten dirty with insane amount of politics and hypocrisy!!

P.S. - Even I have observed a glaring contradiction between the various stands you (Mr. Barua) have assumed as and when you found them convenient, but I will save that story for a later day!!

Unknown said...

Dear Mr. Barua,

First, some background on myself:

1. I passed my 12th standard exams from the Gujarat Board.

2. I studied in a fairly typical school: Not the best in my city, but not very bad either. My attendance record was decent.

3. I took coaching for my board exams throughout standard 11 and standard 12.

4. I took coaching for JEE for about nine months, when I was in my 12th standard.

I write all this to convince you that I am in a unique position to comment on this issue (because I took board and JEE coaching, and studied diligently in school simultaneously). Here are my opinions regarding the issue:

1. Will the new pattern reduce stress on the students?

I think that the new pattern will increase stress, rather than reducing it. The reason students are stressed out today is because of unfair expectations from society, not because of the large number of exams. Reducing the number of exams will only serve to concentrate all the stress on that one day, and in fact create tension in the minds of students (because failure to perform on one day will suddenly have far worse consequences).

2. Do board marks test a student's calibre?
I can only speak for the Gujarat Board. In my experience, board marks have very little correlation with a student's calibre. I agree that JEE is far from perfect, yet if we use a standardised measure to test student intelligence - for example, performance in the national Olympiads or in the KVPY written test - the students who are able to do well in these standardised tests do far better in JEE than in the boards. Prof. Sanghi, if the lack of data is an issue, I have suggested a readily available dataset. Secondly, in Gujarat, board exams are designed not to test your knowledge of the subject, but to test your knowledge of the textbooks. In the Gujarat boards, most questions (certainly more than 90%) are directly copied-and-pasted from the textbook. Many of the remaining 10% are also minor modifications (such as changing a number here and there or changing the ordering of the options in MCQs).

3. Is it fair to normalise marks across boards according to percentile score?
The tendency in Gujarat is the following: Motivated and well-informed students tend to join CBSE. (Of course there are exceptions, but this is the tendency on an average.) Secondly, CBSE schools tend to be located in urban or semi-urban neighbourhoods, while the vast majority of schools in rural areas are affiliated with the state board. As a result, it is far easier to score a higher percentile in the state board than in CBSE. Hence the proposed system is unfair to CBSE students.

4. Will this lessen the monopoly of coaching institutes?
As I have already stated, I am from an average school. I remember that every single student in my class except one took coaching for the boards. (The one student who did not take coaching probably did so because of monetary reasons.) If the proposed system is implemented, the existing coaching classes will simply shift their focus. This will not reduce the influence of the coaching classes in any way.

Since only 4096 characters are accepted, I will post the rest in another comment.

Unknown said...

5. How do we reduce the influence of the coaching classes?
Complaining about coaching classes does not serve any purpose. The only reason students take coaching is because they are not satisfied with their schoolteachers. (I remember my English teacher pronouncing 'fury' as 'furry'.) Most schoolteachers from the state board are incapable of explaining the concepts as well as teachers from coaching classes. Reducing everyone down to the lowest common denominator is counter-productive. Mr. Barua (and other directors), you were asking for constructive suggestions, so here is one: If the government wants to lessen the influence of coaching classes, the only correct way is to train schoolteachers sufficiently well and to provide them sufficient incentives to teach as well as coaching classes.

Let me try to understand the MHRD's decision here: Students face stress because of coaching classes (wrong), so we want to reduce the monopoly of coaching classes. Giving weight to the board marks will reduce the monopoly (wrong) and will test a student's concepts (wrong). A reasonable normalisation across boards also exists (wrong again). Not one argument makes sense!

Anonymous said...

@Ankur June 10, 2012 10:17 PM:

If you mean your comment at June 7, 2012 2:49 AM on the other thread:

Your analysis is in terms of absolute marks. What matters are the percentiles. I think Prof. Barua adequately pointed this part out; see his comment of June 10, 2012 6:33 PM on this thread.

BTW, your analysis refers to the population average. Please see the discussion below.

@Piyush June 11, 2012 12:24 AM and
@Ankur June 10, 2012 10:17 PM:

What is the relevance of Central Limit Theorem here? CLT applies to the normal distribution. Empirically, however, test scores at the higher end (the RHS tail) usually don't conform to the normal distribution. That anyway doesn't matter for the boards to be comparable via percentiles.

What matters is this: Suppose you plot percentage (or fraction) of students at a given test score f(s) vs. test scores out of 100 (s), for the Pune board on a plastic transparency, and produce a similar graph for the CBSE board on another transparency, and then superpose these two graphs, they should coincide, of course, within statistical limits. That's what matters.

If they do coincide, then the two boards indeed are comparable via *percentiles*.

Notice, this is a rather empirical approach. We don't care for many things.

(i) We don't care whether the random variable (RV) here (viz. individual student's ability/score) is identically distributed or not---in reality, it isn't. We don't care whether the distribution for sum of a large number of RVs therefore comes out normal or not. At tails, it usually doesn't.

(ii) And, since the data here are normalized (by taking the fraction of students, and by normalization to the same maximum marks of 100), we primarily don't care for the population size either. Size enters analysis only because of the relatively higher noise at smaller sample sizes.

(iii) Piyush, if the shapes of the normalized graphs match, it of course means that their variances match. In fact the requirement of shape matching is far more stringent. Variance is a single and simple measure. Just like with averages, it is very easy to think of distributions of identical variances but of differing shapes. In contrast, fixing the shape automatically fixes the variance.

Finally, notice, for higher-end admissions, the shape of the distribution at the lower-end (say lower 70%, 80%, in the cumulative distribution graph, that is) hardly matters. Thus, the shapes don't have to match at the lower-end.

The whole analysis rests on a premise that basically asks you believe in the metaphorical (Hindi) "judawaa bhaai" case. Two identical twins (who have a very high probability of identical intelligence) are separated at birth. One is adopted by an IIT Kanpur professor and attends the CBSE/ICSE school; board size 2 lakhs. Another grows up in a rural town in Maharashtra; board size 10 lakhs. The premise says that with similar study hours spent within their boards, given their identical innate abilities, if the IIT Kanpur twin can be statistically expected to have a CBSE rank of 200, then the rural Marathi twin can be expected to have a rank of 1000. That's what the premise essentially means. The rest of it all is just mathematical rearrangements.

I think the premise is fair---assuming that neither board cheats in the lab exams etc.---which is a worrisome consideration. Therefore, though theoretically board percentiles are comparable, practically, in my solution (see my blog), I use the boards only as a filter. (And not because I think that percentiles by boards are worse than those by the JEE.)

Phew! Such a long comment. Give me (too) a break, please!

Ajit
[E&OE]

Unknown said...

Marks across various boards (in any format) can be normalized or not?? this question should only come after when we are very sure that board exams are fair (I mean corruption free) and the toppers of board exams are the "Real Toppers".
For example I am from Rajasthan board and in my time Phy, chem and maths were each 150 marks. 90 theory, 10 sessional and 50 practical. so these 60 marks were completely random component in the total marks. there was a lot of biasness.
Fortunately I secured 17th rank in my board in the state but was unable to crack JEE as i didn't attend coaching classes.
and in coaching classes I felt the difference between board and JEE.

so how come these normalization process eliminate these kind of toppers. NO WAY.......

aaa said...

why can't our dear minister shri sibalji just announce that the implementation is postponed to 2014 and get onto discussion on the issue. At least, it will clear the issue for current Standard 12 students and students who will be attempting their second chance. How the proposed normalisation of marks across various state boards and cbse etc can be tried out in 2013 as a sample procedure and then compared with actual results. And if sibal ji wants class 12 boards to be given attention to, for the year 2013, he can increase the cutoff to 70% instead of 60%. What is the big issue about implementing this untried and untested procedure with effect from 2013 and create confusion all over in the minds of students, colleges, academia, parents, schools, IITs, and what not?

Anonymous said...

To sibbal and other directors of IIT,
Here is how your so called dream of removing coaching is playing out
"http://www.fiitjee.com/fiitjeejava/jsp/homepagenew.jsp?centcode=55&centname=Bangalore"
This is an indication of how system will be recreated to favor elite from metro.

Ankur Kulkarni said...

ajitjadhav: "Suppose you plot percentage (or fraction) of students at a given test score f(s) vs. test scores out of 100 (s), for the Pune board on a plastic transparency, and produce a similar graph for the CBSE board on another transparency, and then superpose these two graphs, they should coincide, of course, within statistical limits."

Why? I was not able to show this. Why don't you prove it for me and we can see what assumptions you require. Also make clear what you are assuming as the goal of the selection procedure and how it can be achieved by equating percentiles on the above distributions.

aaa said...

I find that the discussion on the subject is centred around how the marks will be normalised or how board exam results should be taken into consideration/not taken into consideration. Why there is no discussion as to what is the great hurry to implement the same with effect from 2013 which is in the middle of session for student in Class 12 who are the worst affected lot. Why cannot we pursue the case that immediately it should be announced that the new system will be implemented from 2014 and have healthy discussion on the subject. I REQUEST ALL TO RAISE THE QUESTION 'WHAT IS THE BIG HURRY TO IMPLEMENT IT FROM 2013 AND PLAY WITH THE LIVES OF LACS OF STUDENTS?
Mr Sanghi I request you as well to bring out this issue forcefully .

Dheeraj Sanghi said...

@aaa, I am traveling with lot of agenda and intermittent access to Internet. Will be back to blogging on Thursday.

An Indian said...

Even the village nanny is educated enough to appreciate that the bad habits do not go just because responsibility is thrust upon http://www.youtube.com/watch?v=vwjCsEh9Y1k

Besides representation to PM one may consider taking Mamata Banerjee into confidence for roll back who may get it right quickly with less amount of bloodshed.

An Indian said...

Even a village nanny appreciates that the things do not improve just because responsibility is thrust upon somebody
http://www.youtube.com/watch?v=vwjCsEh9Y1k

Mamata Banerjee should be approached for roll back if one wants least amount of bloodshed

Abhishek Gaurav said...

The class XII board examination is made asses the quality of all students i.e it tries to measure everyone and on the way compromises the fine resolution. On the other hand JEE does not care about 95% of students but tries hard to separate 99.98th percentile student from 99.99th percentile student. Don't see a good way to measure these two ..

Anonymous said...

@Ankur June 11, 2012 10:14 PM

Answer in two parts. Part 1/2.

>>"Why? I was not able to show this."

To repeat, that's because your analysis is in terms of marks, not percentiles. Shake off that mode of thought completely off for a moment, and have a fresh look at what Prof. Barua, ISI Kolkata people, and I have been saying. Please do give it the time it requires.

>>"Why don't you prove it for me and we can see what assumptions you require."

Take a series of the "juDawaa bhaai" cases, with pairs of identical twins having their respective native intellectual abilities running from the levels of morons (0 score) to geniuses (100 score), also covering all the scores in between. For any given pair of the twins from such an entire series, *if* *each* pair scores identical percentile scores in their respective boards, then the two normalized graphs for the two boards would be coincident. Do you agree with this assertion? If so, the rest is easy. If not, take time to let that part sink in.

Now, since pairs do exist at every level of performance (from 0 to 100), completeness is ensured thereby, and we may therefore take the converse. Do just that. As the converse: *If* the graphs do actually coincide, *then* scorers (or people-pairs) having the same given level of a percentile score must be taken as "identical twins" in a certain sense---viz., they must be taken equal in their native ability. Due to completeness, the converse is forceful; it says "must." Agreed with the converse?

That's all there is to it.

You had said in the other thread:

>> "We need to assume that the average of the abilities of all students is the same for every board."

To repeat, your analysis is in terms of the average marks. My (and Prof. Barua's and ISI Kolkata's) analysis is in terms of the identity of shape of the normalized distribution function---i.e., identity of the percentile scores for *each* pair taken one at a time, for the entire series them. Just read what I am saying more carefully, perhaps taking concrete hypothetical examples, drawing hypothetical graphs, and the points I make should be easily apparent. If not, I am sorry I cannot help you any further in a forum like this; please solicit help from a local statistician (including asking him to point out any contradictions/errors in my line of reasoning.)

Ajit
[E&OE]

Anonymous said...

@Ankur June 11, 2012 10:14 PM

Answer in two parts. Part 1 of 2.

>>"Why? I was not able to show this."

To repeat, that's because your analysis is in terms of marks, not percentiles. Shake off that mode of thought completely off for a moment, and have a fresh look at what Prof. Barua, ISI Kolkata people, and I have been saying. Please do give it the time it requires.

>>"Why don't you prove it for me and we can see what assumptions you require."

Take a series of the "juDawaa bhaai" cases, with pairs of identical twins having their respective native intellectual abilities running from the levels of morons (0 score) to geniuses (100 score), also covering all the scores in between. For any given pair of the twins from such an entire series, *if* *each* pair scores identical percentile scores in their respective boards, then the two normalized graphs for the two boards would be coincident. Do you agree with this assertion? If so, the rest is easy. If not, take time to let that part sink in.

Now, since pairs do exist at every level of performance (from 0 to 100), completeness is ensured thereby, and we may therefore take the converse. Do just that. As the converse: *If* the graphs do actually coincide, *then* scorers (or people-pairs) having the same given level of a percentile score must be taken as "identical twins" in a certain sense---viz., they must be taken equal in their native ability. Due to completeness, the converse is forceful; it says "must." Agreed with the converse?

That's all there is to it.

You had said in the other thread:

>> "We need to assume that the average of the abilities of all students is the same for every board."

To repeat, your analysis is in terms of the average marks. My (and Prof. Barua's and ISI Kolkata's) analysis is in terms of the identity of shape of the normalized distribution function---i.e., identity of the percentile scores for *each* pair taken one at a time, for the entire series them. Just read what I am saying more carefully, perhaps taking concrete hypothetical examples, drawing hypothetical graphs, and the points I make should be easily apparent. If not, I am sorry I cannot help you any further in a forum like this; please solicit help from a local statistician (including asking him to point out any contradictions/errors in my line of reasoning.)

Ajit
[E&OE]

Anonymous said...

@Ankur June 11, 2012 10:14 PM

Answer in two parts. Part 2 of 2.

>>"Also make clear what you are assuming as the goal of the selection procedure and how it can be achieved by equating percentiles on the above distributions."

Without having to determine the absolute values of native abilities, the selection procedure should ensure two things: (i) Parity: If a person of a certain ability from board A gets selected at a certain global rank, then another person of the same native ability but from board B should also get selected at the same rank, and (ii) Proper Orderliness: If a person of relatively lower native ability gets selected at a certain rank, then another person of a relatively higher native ability must also get selected with a higher rank.

For the reasons spelt so far, *if* the empirical graphs do coincide, *then* the respective percentile ranks would indeed allow you to ensure this kind of a parity and proper orderliness in the selection procedure.

Notice, we don't care determining the absolute value of the native ability---all that we are concerned about is exploiting the relative measures. We use the pigeon-hole logic to make sure that both of the twins coming from of every pair is selected in, starting in order from the most talented pair and then going down in merit, until all available seats all filled.

Notice, the pigeon-holing principle *can* hold even if board A tests them on mathematics and board B tests them on modern art or rapping. So long as the distribution functions happen to be identical for the orderly pairs of twins, the pigeon-holing is effective. The nature of the two board exams does not matter at all. Now, it's a different matter that the differences between the boards in India are not as great as that in mathematics and rapping. Nor, for that matter, the difference between the boards and the JEE (regardless of what JPBTIs are fond of thinking about the issue, that's the reality).

One final point. This same analysis *can* hold if, instead of a series of identical twins in different boards, you take the scores of the same set of students but on two different examinations, e.g., the JEE, and, four years later, the GATE or the subject GRE or the general GRE. Asking for parity and proper orderliness, in this case, simply means saying that the JEE *percentiles* and the general GRE *percentiles* (not scores) correlation is 100%. That's what it means. I expanded the discussion a bit, but only to show how its other application should be naturally acceptable to you.

I think as far as this mathematics and proving and all that business is concerned, I am done. Look elsewhere if still not convinced.

Ajit
[E&OE]

Ankur Kulkarni said...

ajitjadhav: this far too vague to be convincing. How are the graphs obtained, how are you normalizing? Besides, I still don't know what assumptions you have made.

I did several analyses before writing up the one you read; the cleanest way I found of dealing with the problem was to go by marks not by percentiles or order statistics.

I sense a patronizing tone in your answer. I won't respond further if it continues.

An Indian said...

Hope good sense will prevail. Pls. go through the media publication pasted below. As I told before, the amount of bloodshed will be less by this - each drop of blood of an academician makes this land less honourable, a land where academicians were seen in the highest esteem, kings touched their feet, bowed down before them ... allow them to do their job and we remain grateful to them for the job they are doing.

http://in.news.yahoo.com/iit-kharagpur-teachers-look-mamata-side-183000400.html

Kolkata, June 12 -- The teachers' association of Indian Institute of Technology, Kharagpur, has decided to take its fight against the Centre's single entrance test move to West Bengal chief minister Mamata Banerjee.

"The association has decided to write to the CM and seek her intervention," said a faculty member of IIT-KGP. "Introducing the single entrance test for all the engineering colleges in the country from 2013 will ruin the entire IIT system."

The state government appeared ready to give the teachers a patient hearing. "Let the teachers write to us. Then we will decide," West Bengal education minister Bratya Basu said on Monday.

Bengal has already objected to the human resource development ministry's single test formula, which recommends giving weight to marks secured in Class 12 board examinations. Securing high marks in the Bengal board examination is perceived to be more difficult in comparison to other boards.

The All India IIT Faculty Federation welcomed the IITKGP decision to write to Banerjee. "She has a very strong personality. If IIT-KGP teachers can make her understand the problem, we can expect her to take up the matter with the Centre," said Atul Mittal, general secretary of the federation.

boris said...

Firstly great step by IIT-K.

I was highly impressed by Dr. Sanghi and the Prof of IIT-D.Mohandas Pai was not logical.

A few points:

1)We can't replicate the Chinese system as Mr. Pai wants they are a communist govt under one party rule since the inception of the PRC,their model can work only in countries like china and the former USSR.

2)Why should we build 100 IIT's apart from financial issues and infrastructure problems that will arise,why should IIT's bear the entire burden of education??is it not the duty of private institutes to give quality education at the same level rather than throw in donation based seats.

3)Coaching institutes will stay no matter what the pattern,since 2000 the pattern of the JEE has changed time and again but Coaching has only blossomed be it a hard or an easy exam.As it is ISEET batches have started at coaching institutes.

4)School Education will not improve by pointing fingers at the IIT,this problem has to be solved by the CBSE and state boards for having curriculum and exams that encourage rote learning.IIT since its inception is doing its job of imparting world class technical education,school education is not their domain.

5)One exam for all is suicide,you cant expect someone who just wants admission to his state engg. college to attempt something as hard as the JEE rather than the easier state entrance that his seniors gave.The JEE difficulty is modeled to select exceptional candidates.

boris said...

Another point is why is there an age limit of 25 to pursue engg at IIT's and NIT's,in most countries we have 40 year old's pursuing Bachelors in Engineering.Age limits work for govt. service and military not in education.

Anonymous said...

@Ankur June 12, 2012 9:44 AM:

"I don't understand it, so you must be vague. And, let me get away thus belittling you, counting on my JPBTI status, UIUC affiliation, etc."

Or

"I don't have the attention span to hold all the points you painstakingly mention in your multiple responses, so what you say must be vague. And, I can still get away by writing just one-two liner comments. Ultimately, I've got some kind of a right on your time."

Yeah, right. That's not being snobbish/patronizing. Expecting detailed explanations of every small dumb point from the next person, but writing only one/two liners when it comes to one's own response, without taking the responsibility even to jot down precisely what part one *has* understood. No, that's not being Brahministical, that's not showing a JPBTI's snobbishness, that's not being patronizing.

My final advice: Ankur (and I do know your name), do get in touch with a local statistician if he knows what he talks about---though I expected you to know better, with your PhD in S&T field and your post-doc under a statistician, not to mention your occasional comments at other blogs expressing at least part admiration for Ayn Rand's ideas. I now certainly have corrected my expectations.

I won't pursue replying you back, unless I have to. And make doubly sure that you are not being deliberately dumb or snobbish or Brahministical or patronizing in your reply, if you at all reply---is that clear?

Ajit
[E&OE]

Piyush said...

ajitrjadhav:

The curves cannot be expected to match up until the variances do, too. For from any analysis of the curves, no one seems to be willing to put forward any data that that even the variances match up.

Unknown said...

@ajitjadhav:

I address this question to you because you seem to be the only person from the other side interested in rational debate.

Would you agree that: 1) The national Olympiads, and 2) The KVPY written test, are genuine tests of a student's understanding of concepts?

If you do agree with this, would you agree that a statistical analysis can be performed as follows: Measure the correlation between a student's performance in these exams, and his performance in: 1)JEE 2)AIEEE 3)XII Boards. Would you agree that the best exam out of these three is the one showing the best correlation? Would you further agree that, if the Boards show a very weak correlation (or a negative one, depending on your method), it is a very bad idea to consider the Boards as a measure of innate ability OR acquired ability OR some combination?

If you agree, carry out the above analysis. I think the results will surprise you.

Manish Verma said...

Some graphs are shown at http://www.cse.iitk.ac.in/users/dheeraj/jee/ramasami-ann6-ppt.pdf by Indian Statistical Institute (ISI), Kolkata (http://www.isical.ac.in/). The data sources used and assumptions made are mentioned at http://www.cse.iitk.ac.in/users/dheeraj/jee/ramasami-ann6.pdf . Since it is stated that analysis is based on the 4 year data for 4 boards, the data should be available with them, while it need not necessarily be there in the public domain at this point. There is a good possibility that the data should be available for other boards as well with respective board. If an independent body wants all that info, it could demand lot of involvement and sometimes even the use of RTI. Moreover, such type of analysis needs to be done by an authority. May be ISI folks need to jump into the discussion.

chitta said...
This comment has been removed by the author.
aaa said...

ajijadhav ji,
u seem to be an expert in technical issues or so u think. why it is not important for u to ask IIT directors (who are supporting) and ministry of HRD, as to why their database is not in the open for discussion on the issue of normalisation. Why cannot it be shared with the public.

Are u not aware that the proposed system on 28 may 2012 was never in the public domain in the present form. What i mean to say there were many combinations which were suggested but the combination which has been approved was not in the public domain in this manner. So, would it not be fair on anyone's part to put this finalised system in public domain and ask for views of common public, students, parents, academia and IITs,NITs etc.
Moreover, what is the rationale of insisting on implementing it from 2013. Why not 2014. I am sure by then with your analysis, u would be able to prove that the statistical method is alright.

Anonymous said...

@ Tanmay Mudholkar June 12, 2012 11:55 AM.

Three-part answer: Part 1 of 3.

Apart from @kingkhan (and of course Prof. Barua et al.) it was your comments that I have appreciated.

All tests aim to measure a student's knowledge, whether national Olympiads, KVPY, JEE---or AIEEE, or XII Boards. But they end up measuring different aspects of knowledge to differing degrees. Loosely speaking, some of them are more difficult to answer than others. More exactly speaking, the "more difficult" tests are more sensitive (i.e. they carry better resolutions) at the higher end, as compared to the other tests. NOlymp, KVPY and JEE all are among these exams.

However, do notice, sensitivity (at the higher end) doesn't at all mean accuracy (even at just at that higher end). For that matter, it doesn't even mean precision. As you would know, these are entirely different concepts.

Every test carries some:

(i) inaccuracy ("the n-th ranker should have been the topper, really"; "the first-ranker should have got 96% and not 95.5% marks"; etc.),

(ii) insensitivity ("Ram and Sham both got exactly 75.00 percentile even though Sham is slightly better of the two"),

(iii) imprecision ("there is a band (a distribution) in which an individual's score lies on repeated attempts but with the same level of preparation"; "the correlation between scores of a group of students on two JEE practice tests, even if designed by the same IIT professors, is not 100%"), etc.

In view of test attributes like these (call them limitations if you wish), any test provides at best only an estimate of your knowledge/skills.

That's one of the reasons why a single test must not be made a criterion for too many things such as specific IIT campus and branch decisions etc.

>> "Would you agree that: 1) The national Olympiads ... Would you agree that the best exam out of these three is the one showing the best correlation?"

The difficulty here is this: If I (or anyone else) cannot even properly define any standard to measure the absolute accuracy of a given test, how can I say that some one of them (say the Olympiads/KVPY/JEE) is a "genuine" one (meaning: most accurate)? I could have called them more "genuine" if I could have been assured that their inaccuracy was zero or at least negligible. But, before I could assert that, I would have to know (and be able to clearly state) the standard by which I was measuring that accuracy. Mind you, such a standard, to be applicable, would itself involve actual testing via yet another test. But no such a hypothetical test giving 100% accuracy exists.

The reason for that, in essence, is this. For every valid thing that such a test could be shown to measure satisfactorily, people can easily come up with many other valid things that should have been measured by that test but which are not. The overall problem lies right there.

Ajit
[E&OE]

Anonymous said...

@ Tanmay Mudholkar June 12, 2012 11:55 AM.

Three-part answer: Part 2 of 3.

Knowledge is multi-dimensional, transiently changing, and requires application for its grasp to be even estimated let alone measured very accurately.

Since knowledge held by a student is multi-dimensional, single tests do not even statistically probe all its dimensions. ("Sir, I had studied that topic so well, but there was no question on it in the exam. Could I still look forward to an A now?") Any test result---even that of JEE etc.---is like asking to accept a square or a circle as the shape of something when the actual shape of that thing is rather like an ink-bloat or an amoeba. (And, this is just a 2D analogy.)

Knowledge is transient, and requires application in the act of measuring it. Your preparation may remain the same, and two tests may probe more or less the same topic from more or less the same angle that you had studied them, but if you temporarily lose focus while answering the second test, you degrade your score in it.

Since knowledge is multi-dimensional, transient and requiring application again and again, reliably measuring the state of knowledge becomes a very complex problem. Fortunately, we are not required to solve this problem with 100% accuracy, as far as the admission procedures go.

No educational institution is concerned with setting up admission bars in terms of absolute knowledge---not to as detailed a degree as consideration as the competitive intakes require. When it comes to admission decisions, the bars always are in terms of relative performances. Selecting 10k out of 10 lakhs is the problem---not determining whether the topper among them still knows enough physics on this, this, this and that count, to this, this, this, and that degree, or not. No IIT throws out the JEE topper simply because he failed to crack the more difficult 30% of the problems (which, incidentally, the usual statistic). The working assumption here is: if he doesn't know enough physics, he will pick it up---he is the best of the available lot. If so, if you can be so generous when it comes to the absolute knowledge level of the student, why not apply similar considerations when it comes to tests as well.

So, what we really need aren't tests that are "genuine" i.e. 100% accurate in probing all aspects of knowledge. What we need are the tests that can *reliably* pick out top 10k from 10 lakhs.

Keeping aside factors like possible malpractices in the practical/oral/internals section, or deliberate malpractices in grading and moderating the results, the XII board exams could easily perform this task---provided that branches were not to be allocated right at the time of admissions. The whole business of AIR hinges on that requirement---which is entirely unnecessary.

Boards, too, are exams. They, too, test something. It's just that they are easier---or, more accurately, their resolution at the higher-end is not very good. So, allocating branches on their basis would be a poor decision, overall unjust to students. (However, the difference in resolution, IMO, is not as great as JPBTIs are prone to make it out.)


Ajit
[E&OE]

Anonymous said...

@ Tanmay Mudholkar June 12, 2012 11:55 AM.

Three-part answer: Part 3 of 3.

Or, is it your contention that someone who can top JEE will fail to be within top 10k if you go by the Boards route? Notice, if branches would not be decided at the time admissions, if AIR ranks were not to be determined at all, but only top 10k were to be filtered in, then our burden would reduce so much that the only things we would want to be concerned with would be simple things like these: asking if someone who should have been the 9995th guy and someone who should have been the topper guy, have both indeed made it to our 10k list or not. (And, how do we know if somone should have been 9995th/topper? That's just a probabilistic projection, based on consistency of performance in an array of hypothetical tests. That's all it means.)

On the other hand, if you wish to say that it is possible for a JEE top ranker to fail to be within top 10k in the Boards route, then I would say that his knowledge is far too one-dimensional for him to be pursuing normal courses of science and engineering, whether at COEP or at IIT---and that JEE is far too bad an exam to decide overall IIT admissions anyway.

But then, I am also known to make a lot of dramatic, even "unbelievable" statements. The JEE topper for my batch was, I suppose and vaguely recollect (but never have cared to get to know who precisely it was): Dr. Joy Thomas. Now, Dr. Thomas has made many fine achievements. No disrespect intended, and I wish to take nothing away from him, but still, frankly, I don't think that the level between his actual achievements over the past 20-30 years, and those of many others, now in retrospect (or "mid-o-spect") turns out to be all that dramatic. Indeed, arguments could conceivably be forwarded that certain others might have actually achieved superior intellectual achievements than him (though I won't identify who :) ). I personally know of two/three President's Gold Medal-winning JPBTIs, and, after 10--30 years, they still don't impress me as much as many others actually do. (And, to be fair to them, not a single one of them has any issue accepting the factual content underlying that statement---it seems that toppers themselves approach me a bit more reasonably than those who put them on pedestals). Some/many of these others of superior achievements have attended colleges other than IITs, and they have followed the XII boards route. Point is: the XII boards route was good enough to pick them up. For instance, Venky (Nobel laureate), Mukesh Ambani and R. Mashelkar (UDCT), S Patankar of CFD (COEP), et al. Keep that part in mind.

Though, as I said, resolution/sensitivity of today's XII boards across all states doesn't seem to be as good as one would like to see. And, then, there of course are those malpractices considerations---a very very serious matter.

Finally, as mentioned on the LinkedIn thread (about 6+ months ago (BTW, I assume you have seen my personal blog post)) I would love to see specific research documents showing degree of correlations for percentile performances between: (i) JEE and Boards, (ii) JEE and CGPAs, (iii) JEE & general GRE, (iii) JEE and subject GRE/GATE, and then, of course, I would be also delighted to see similar correlations between JEE and KVPY/National Talent Search Examination percentiles/State scholarship percentiles, etc. And, I would love to see correlations involving Boards in place of JEE for all of the above. The most important of them, for obvious reasons, would be, I think: JEE vs. general GRE and XII Boards vs. general GRE.

@aaa I join you in requesting all the authorities, including the minister and the IIT directors, to make research and data involving the above-mentioned correlations publicly available.

Ajit
[E&OE]

Shishir said...

@Ajit, I agree that the Board exams could have served the purpose.
And indeed, they did serve the purpose in my time . Nearly all Engg colleges including RECs (old avatars of NITs)took admission on the basis of Intermediate or Higher Secondary marks. Yes,Engineers of that era did reasonably well too, nationally or Internationally.(IITs also took a few board toppers directly!)
But with time, something snapped or rusted away. Nearly all the boards started producing high marks students in large numbers. So large that colleges found difficult to devise a criteria to distinguish between them for admission (Imagine, half of students in a batch passing out with 9 or higher CPI in IIT/K).
So, they had to resort to entrance exams. (Please remember that one of the objective of any test is to winnow the chaff from grain. If a test can't separate out excellent, good , satisfactory and the rest, it is not a very well designed test.In your parlance, the exam result marks should be distributed normally for those who are in competition zone.)
In order to build a credibility of the test, the test results need to be validated by the performance of the final product.
It is here the JEE has excelled. Year after, JEE high rankers have performed well in IIT and thereafter. It has been extremely fair for nearly half a century of its existence. Something worth being weighed in Gold , given the state of our environment.
Why tinker with it?
There could be flaws in JEE, but they have to corrected through evolutionary process as JEE has proved to be the best bet we have at present and makers of JEE are faculty members , to a large extent.
Now, the technical part of your argument. Your arguments could have been valid if state boards (including CBSE and ICSE) produce students of consistent quality. Though no data is available either for or against it, anecdotal evidence is that the quality is not consistent. This leads to comparison of marks pattern (whichever theory you use)meaningless, as consistency of quality of the product is the necessary condition of any statistical comparison of patterns.

Sriram said...

The world over, entrance criteria focus on the preparedness of the student in particular domains. In the USA, competitive students not only do the SAT but also take a variety of advanced placement (AP) tests that assess domain-specific knowledge/skills in Physics, Computer Science etc. These subject tests have increased in importance, in part, because of the lack of comparability of high school grades. Countries like Japan, Korea, and China conduct rigorous college entrance examinations to assess preparedness. If mere aptitude was sufficient, they would have used a general aptitude test like the SAT. Subject knowledge matters.

In the present context, HRD proposes that (i) performance on high school board examinations are a valid index of preparedness for IIT education, and (ii) they have incremental validity, over and above the JEE examination alone. Using high school exam performance as a criterion is complicated by the presence of numerous board exams. Discussions about methods of incorporating board exam performance (raw scores, percentiles or other transformations, plotting/normalizing distributions etc.) presuppose the “comparability assumption”. The numbers of students taking different board exams vary widely (e.g., from a low of 20,000 to a high of 20 lakhs, two orders of magnitude). Are the top 5% students in each board are comparable in preparedness in that they have similar aptitude AND ability in math and science? If this is true, we should conclude that the top 1000 students in the ICSE board are comparable to the top 20,000 students in the UP board. Are such conclusions warranted? Are the top 2000 UP-board rankers similar to the top 100 ICSE rankers, in terms of preparedness for IIT education?

Under what circumstances must the comparability assumption be necessarily true? Only if it is presumed that the students are randomly assigned to the boards, as in a controlled experiment. It is apparent that random assignment does NOT hold. Students are either self-selected into various boards or assigned by systematic factors (e.g., geographical location). Once in a particular board system and school, they are under the influence of other systematic factors (quality of teachers, curriculum, peers, coaching ). The influence of these factors on students preparedness make a direct comparison of performance on different board exams invalid. Arguing the fine details (raw scores vs. percentiles vs. other transforms) begs the question of whether the comparability assumption is valid in the first place.

To check the comparability assumption we can examine the distribution of scores on entrance exams and look at the covariation with board examination performance. If we compute the correlation between the JEE/AIEE and the respective board exams. There would be 30 odd correlations and the average correlation would indicate the degree of validity of board exam performance. Significant differences in this correlation, across different board exams, would indicate that not all board exams are created equal because stronger correlations would be associated with board exams that are more valid.

More importantly, if it is found that students from certain boards do better on average in JEE/AIEEE than others, it suggests they are better prepared and consequently, the use of percentiles is, ipso facto, strongly biased against them. Without asking the right questions and verifying assumptions at each stage of the analysis in a honest and transparent manner, we will end up with nonsensical conclusions and unfair policies that, long term, will hurt excellence in technical education.

The focus should really be on methods to improve the JEE by posing questions such as: Which JEE components are more valid? Should computer science and general knowledge be included? Is a 2-stage selection process better? Should open-ended problems be included? Ask the right questions and focus on what really matters.

Unknown said...

PART 1 OF 2

@ajitjadhav: I agree fully with @Shishir above. I was about to reply to your post, but he seems to have covered a lot of the points himself.

Firstly, I thank you for the time and effort you seem to have put in just on the comments section of a blog. I have some points regarding your reply to my query. (I have slightly changed the order of your points.)

"All tests ... end up measuring different aspects of knowledge to differing degrees. Loosely speaking, some of them are more difficult to answer than others."

Notice that the second statement has nothing to do with the first! I fully agree with the first statement. There are indeed different aspects to knowledge, and different exams aim at testing different attributes. The question is: What attributes should any good engineering college be looking for? I can still form an easy JEE paper which would look completely different from an easy Board paper. Similarly, a difficult JEE paper is completely different from a difficult Board paper. (The reason Boards do not set difficult papers is because they want to pass enough people.) I cannot emphasise this difference enough. What sets JEE apart from other qualifying exams is not just its difficulty, although that is a part - it is the principle that conceptual understanding is the only thing which must be rewarded. Would you not agree that rote learning plays no role in any higher education? I am not being elitist here; I think all qualifying exams should test conceptual understanding. They can be easier versions of JEE! Memory does play a role in education, but only insofar as it helps the student to assimilate different concepts. The only way of testing this memory is conceptual questions!

"In view of test attributes like these (call them limitations if you wish), any test provides at best only an estimate of your knowledge/skills."

I agree completely! In fact, I would be the first person to say that JEE is not perfect. However, there is such a thing as good estimates and bad estimates. The goodness of an estimate, being an ordinal measure and not a cardinal one, cannot be measured numerically without ambiguity; and my point is that JEE is far, far better an exam than any Board!

"The working assumption here is: if he doesn't know enough physics, he will pick it up---he is the best of the available lot. If so, if you can be so generous when it comes to the absolute knowledge level of the student, why not apply similar considerations when it comes to tests as well."

I am sorry, Sir, but nobody is claiming that JEE 1 knows physics absolutely! The only claim is that, to the best possible accuracy, JEE N should be, on an average, better at solving conceptual problems (as JEE problems are) than JEE N+1. This assumption is NOT valid for the Boards!

Let me continue this argument in another post.

Unknown said...

PART 2 OF 2

"The difficulty here is this: If I (or anyone else) cannot even properly define any standard to measure the absolute accuracy of a given test, how can I say that some one of them (say the Olympiads/KVPY/JEE) is a "genuine" one (meaning: most accurate)?"

You cannot! As with several things in life, you can only guess. However, your guesses should have some basis in fact. Some comments here: Please bear in mind that the Olympiads are an internationally accepted measure; Indian students, selected on the basis of the National Olympiads, have routinely won prizes, including Gold medals, at the international level. In order to aid your guess, let me recount some of my experiences during my +2: (I give you my word of honour: None of these is made up. All of these are +2 level regular teachers.)

1) My mathematics teacher claimed that pi, being 22/7, is rational.

2) My chemistry teacher claimed that all seawater is a saturated solution of sodium chloride.

3) My physics teacher claimed that the units of resistivity changed if you changed the shape of the substance.

4) My English teacher pronounced 'fury' as 'furry'.

5) The "Thomson effect" was described in our physics textbook; it was predicted by Sir William Thomson, who later became Lord Kelvin - this nugget of information was also mentioned in the textbook. So one teacher had the brilliant idea to ask in an exam: "Explain the Kelvin effect." (I guess the next question was: State Isaac's first law of motion!)

6) Most Board questions (about 90%) were, and are still, copied-and-pasted from the textbook. Of the remaining 10%, most are minor modifications, such as changing a number here and there or changing the ordering of options in MCQs. Just let this fact explode in your head for a moment: Students actually remember that the answer to so-and-so question is option B! Not just 10 microcoulomb - that is too much to remember - but option B!

I could go on and on, but you get the point. These are the same teachers who will set and correct the Board paper.

After all this, does it not outrage your intelligence when someone claims that the Board exams are a measure of ability? What ability of the student, exactly, are you measuring? Not even the ability to memorise - there is only a small textbook to memorise. Even a monkey will do that, given a year and pushy parents.

It saddens my heart, Sir, to see a rational person arguing for the inclusion of Board marks. Please do not lend credibility to the likes of Sibal (an arts student, by the way, who will probably commit exactly the same errors as I mentioned above if asked anything about science) by pretending that there is a rational basis for including Board marks.

Sriram said...

Here is an idea that the IITs can use to check the respective validities of JEE, AIEEE and board exam scores. Typically, in the first year at IIT students do similar core courses and the GPA at the end of the first year can be used as a proxy of how well a given student has coped with the demands of an IIT education (or slacked off).

A significant number of students probably did both the JEE and AIEE and all students have some form of board scores in the same science and math subjects. Use these independent variables to predict the first year gpa, using multiple regression or regression trees.

The size of the coefficients will provide an index of the respective validities of the constructs measured by JEE, AIEEE and board exams. It is plausible that the board exams may have incremental validity in this restricted range (the analysis is post-selection into IIT) as it might measure persistence to learn materials that are not particularly inspiring (that is, for the same JEE rank, a a person doing better on the board exams is also a person who will spend more time studying and consequently doing better on IIT courses). Likewise, it is plausible that the AIEEE scores (for those who do this exam as well), might measure something of interest.

The more general point is that the IITs have tons of data on their students over many years. Why aren't they mining this and arriving at conclusions? Or maybe someone has looked at all this but there isn't any public report.

gautam barua said...

I am sorry for not being able to respond earlier. I am in Delhi attending a meeting which has nothing to do with the JEE issue. While here in Delhi, I find that I have been personally accused by a number of people.
1. Apparently, the IIT Faculty Federation has indirectly criticised me for airing my views in a TV channel. I have no comment to make on that.
2. One faculty member of IITG (Saswata) has accused me of doctoring the minutes of the Senate meeting in which 90% of the members had other views. a) Saswata did not utter a word in that meeting, b) Saswata joined IITG in Nov 2011, and he is an Asst. Prof. in CSE, c) I have had no allegations of this kind from any other faculty member, d) I agree that the decision in the senate meeting was not unanimous, but a clear majority supported the final decision.
3. There is some accusation of me being in league with coaching institutes. I have replied to this elsewhere. I did attend the prize distribution of the only coaching centre in Guwahati at that time, started by an IITGian of the first batch for a number of years. i felt it was my responsibility to encourage kids from Assam to get into the IIT system. At that time, about 12-15 students were getting into IITs from all over Assam. That number has now increased to 70 or so.
4. I did BTech and MTech from IITB. I was a faculty at IITK for 13 years, and I have been a faculty at IITG for 17 years. I have been the mentor Director of IIT Patna for 1 year. I was officiating Directro NIT Silchar for 2.5 year, and Chairman BOG of NIT Silchar for 9 years. I am proud to be an IITian and an NITian.
I request everyone not get personal, and to stick to issues. Surely I have, as a member of the IIT Council, to air my views. I have nowhere asserted that my views are the views of the faculty of IITG. I of course hope they are.

Gautam Barua

Saswata said...
This comment has been removed by the author.
Anonymous said...

All:

Unless Prof. Sanghi (or the moderator) allows the third part of my 3-part comment to appear "as is" here, I have no interest in this blog.

Someone (possibly including Tanmay/Prof. Barua, if not Prof. Sanghi (!)) please alert me when that happens. I will make sure to check back.

As of now, I regret reposing trust in Prof. Sanghi; didn't save the last edited and submitted version to my hard drive. (It mentioned Dr. Joy Thomas, if that helps you find it, IIT Kanpur CS JPBTI Dean Dheeraj Sanghi.)

IIT Kanpur CS JPBTI Dean Dheeraj, won't repeat that mistake again. Remember that. (You would need to; not I.)

(I *am* making sure to save *this* comment, the last edition of it. Won't mind publishing/forwarding elsewhere, if this, too, is moderated out.)

Ajit
[E&OE]

Sriram said...

Hi Ajit,

Your post part 3 has made it. At the high level, people cannot be compared so easily. Let me assure you that Dr. Thomas who is not in academia has indeed distinguished himself. He is the co-author of a well-known text on Information theory and continues to be deeply admired and respected by his batch mates for many reasons.

I could not help but notice that your blog http://ajitjadhav.wordpress.com/ indicates that you seem to be somewhat unhappy with things and particularly have a deep dislike for people who have studied B.Tech at IITs. A bright guy like you should be recognized and find some happiness and not let himself be defined by other people. Best wishes for the future and hope that you find peace and joy in your life.

Unknown said...

@ajitjadhav: I saw the third part of your comment rather late. It was delivered to me over e-mail.

Some points here:

1) I do not doubt that, in an era when the only students taking coaching were weak students, some schoolteachers were indeed good. Times, however, have changed, and as @shishir pointed out, the very reason JEE was instituted was because Boards were failing to provide a good measure. What happened to the good schoolteachers? I don't know. I think they may have been unintended victims of India's economic development: Suddenly, far more attractive jobs (in terms of salary) were available to those trained in science. Whatever the reason, good schoolteachers are rare today, and that is a fact.

2) The examples of R Mashelkar and S Patankar are irrelevant. (S Patankar would not have been able to take admission into IITs for his Bachelors. They were not even instituted when he began his undergraduate studies. Note, however, that he was one of the first Masters students to pass out of IIT Bombay.) Similarly, JEE was not even present when R Mashelkar did his undergraduate studies. At that time, since the IITs admitted students on the basis of Board marks, the fact that he could not get into IIT speaks against Boards, not against JEE.

3) Your third example of Mukesh Ambani is irrelevant. IITs are not for producing businessmen, and anyway I doubt that Ambani performed well in his 12th standard.

4) Your fourth example of Venkataraman is possibly valid. However, I was unable to verify whether or not he had done well in the Boards either (it is a fact that he was unable to qualify for JEE). The only scrap of information I could obtain was that he had done a pre-science course. Did he do it in place of a higher secondary education? I don't know, so I will reserve comment
here.

5) The analysis you have asked for, i.e. comparing a student's performance after he enters IIT to his performance before he does so, is invalid. If IITs fail to train good students sufficiently well, it is a failure of the IIT system (which can be attributed to a myriad of causes such as social expectations, and so on - that is not the debate here) and not a failure of the entrance examination. Believe me, IITs are taking the lack of interest in engineering students very seriously indeed, but no one thinks the solution is to randomise the entrance.

6) In continuation to point 1, your assumption is that Board exams today test extra dimensions of a student's knowledge, which the JEE does not test. After reading my other post 3-4 posts above this one, please tell me what those dimensions are. They are certainly invisible to me.

A concerned academician said...

Dear Prof. Barua,
Just because Dr. Saswata Sannigrahi has joined in Nov 2011, makes him no less able to air his views on a matter that affects not only the students aspiring IIT JEE or the Professors about to teach them, but the entire academicia of India. The changes that are proposed for only a handful of IIT JEE aspirants are going to affect the entire Secondary and Senior Secondary education in India. While the Government lobby wishes to portray that effect as a positive one, the majority of the rest differ in their views. Its true that the Government and people in power (like yourself) can have the last word. However, if you wish to be DEMOCRATIC (!!)in the matter, you must have the guts to hear what the others have to say. And neither YOU nor any Government has the right to put words into the mouth of others, or force them to agree to your views as you are doing now. You know very well the SPEECHES you deliver at meetings of your institute and the MAILS you write to your FACULTY MEMBERS, with a tone that we seldom use for teenagers who have gone askew! How do you think that all the faculty members of your institute who beg to differ from you will have the courage like Dr. Sannigrahi (qudos to him) to come out in the open to confront you? You have already kept their grave ready, and remind them about that too often. We all can guess what will be the fate of courageous people like Dr. Sannigrahi, but there is a belief that TRUTH will never remain hidden, try however much you may.
You may ask, why then do we not wait and watch? Because we are AFRAID of the darkness that will ensue if we do. We are scared that the youth will suffer much more with these changes, than they do now. The JEE aspirants, other Engineering aspirants will suffer double. Those students who have DREAMS OTHER THAN to be JEE aspirants, but wish to study science, will suffer miseries. They too will be forced to come into the RAT RACE of the +2 exams to get admission in a decent college/university thenceforth.
As far as the IIT Guwahati Senate meeting is concerned, these issues, and many more were raised by someone or the other. So, once an issue was raised in the meeting, people like Dr. Sannigrahi may not have thought about repeating it again. Because we “HAD” belief in the system, atleast in the academicia. Often we blame the “politicians” about the game that they play with us. But rarely do we expect people of high learning and understanding like yourself to behave like them. It breaks our hearts to see the person leading us to behave like the ones who we criticize ever so often.
I still hope your long involvement in academics and your goodwill for students and the academic scenario of the country, will prevail over any other influence that may be acting more strong at this very moment.
~ A concerned academician

aaa said...

it seems from TV discussions, blogs etc that directors of some of IITs, Gawhati, k,pur etc have deliberately changed the minutes of their Senate meetings. Well, does it not call for enquiry and immediate suspension of these persons who are spoiling the name of these great institutions by their cowardly act for some unknown reasons or should we say for some known reasons.

Dr Barua has still not answered the question, as to WHAT IS THE BIG HURRY TO IMPLEMENT IT FROM 2013? I HOPE HE FINDS TIME TO REPLY? AND I HOPE HE DOES NOT GIVE THE VERY BASIC AND ILLOGICAL ANSWER THAT THIS HAS BEEN DISCUSSED SINCE LAST TWO YEARS. DISCUSSION FOR A CERTAIN PERIOD OF TIME THAT TOO IN AN OPAQUE MANNER, DOES NOT MAKE A STUPID ISSUE INTO A GOOD DELIBERATED POINT. IN ANY CASE, THE FINAL PROPOSAL IMPLEMENTED BY MIN OF HRD IS MUCH DIFFERENT FROM WHAT IIT SENATES SUGGESTED SO IT HAS COME IN PUBLIC DOMAIN ONLY ON 28 MAY 2012. AND RIGHTLY SO, IT SHOULD HAVE BEEN DISCUSSED IN OPEN WITH ALL ITS IMPLICATIONS, BEFORE TAKING A CALL ON IT. IN ANY CASE, TO IMPLEMENT A NEW PROPOSAL WEF 2013 IS MOST STUPIDEST IDEA ONE CAN EVEN IMAGINE FROM LEARNED PEOPLE LIKE MR SIBAL AND SO CALLED IIT DIRECTORS (SOME OF THEM)

Anonymous said...

@Tanmay Mudholkar June 12, 2012 11:14 PM:

Re. bad teachers. I am amazed at facts in your pts 1 and 3, though 2, 4, and 5 would be quite common for us too.

Teachers don't matter. Not to the top 0.5% caliber students. Not in the Boards system.

I should know. My I std. teacher was properly mad, taking us out, pointing out at clouds, throwing his "Topi" (cap) in air, and shouting with a slur: (Marathi) "chalaa, chalaa, kaapus baghaa, kaapus baghaa, uchalaa to kaapus." (English: hurry, hurry, see the cotton (in the sky), pick up that cotton.) Such things, he meant seriously, not as a game, on those rare days when he at all came to school. Thus began my education. My 2nd and 3rd standards were spent in aadivasi village where newspapers would arrive 3rd day, and the same schoolteacher would teach 1st to 4th standards in the same room using the same single blackboard at the same time. My 5th--9th were spent in a small town high-school where, to make it easy for students to pass, they wouldn't finish even 50% of syllabus over the entire year, and some teachers indeed were as bad as you mention. (Later on, I had to take efforts simply to learn how to finish an entire portion for exams). Yet, right from that class, without coaching, there came ahead at least five: 1 pvt college BE, 1 UDCT BE, 1 COEP ME + successful manager abroad, 1 NCL PhD chemistry + successful researcher abroad, apart from me. (UDCT & COEP are top in Mah., and NCL, top in India). And, I can cite similar examples for other, even smaller towns producing doctors, engineers, CAs, right in our times when there were so few seats. Teachers really don't matter. Not to the extent they are usually thought to be, and certainly not to the top 2% let alone to the top 0.5%. Not in the Boards system anyway.

And, despite those incompetent teachers also acting as paper evaluators, in our times in Mah., the Boards evaluation was by and large fairly OK (even though the exam wasn't well designed for the top end). How come? There is a lot to be said about the effectiveness of a hierarchical system of evaluations, including evaluator selection, real-time supervised evaluations, two further tiers of moderators, and then the final controllers together with statisticians, with moderators+ having authority to reevaluate.

Re. Gujarat board exams being 90% copy-paste + 10% minor modifications in numericals. If so, it is pathetic. (And, since so many bloggers so badly attack Mr. Sibal, let me ask: What has Mr. Mody been doing about it?). ... Anyway, my statistical judgment tells me that such a poor state of affairs should also certainly get reflected in the distribution function for that board; the shape should be distinctly different. I would love to have a look at it. ... That's the trouble. The data are too many (so many states) and the relevant data are not always easily available.

Overall, it indeed is possible that over the years the boards have diluted their standards. But, unless I have the hard data, I won't pass judgment about exactly how bad the situation is, certainly not going by anecdotal evidence of bright young students. Young IIT students could underestimate the Boards, just the way the elderly BA types could overestimate those. Only hard evidence can be our true friend. I await it being made available/being pointed to.

One final point. Sibal had the grace to say that he does not understand the issue (commenting after the IITK resolution). I have seen his debates on TV, right from the early naughties. My estimate of him is much better than what many bloggers seem to carry. As to the political part, it's easily possible (as some have wondered) that this move is nothing but a states-wise quota in disguise, and calculated with the 2014 elections in mind. Possible. If this is really so, then only Sibal himself won't have as much to do with it. If so, bloggers on your side should think: Would they be equally vociferous if it were a JPBTI Jairam Ramesh in place of the BA Sibal?


Ajit
[E&OE]

Anonymous said...

@Tanmay Mudholkar June 13, 2012 10:41 AM:

My point numbers follow yours.

1. People might have thought it better to institute JEE because Boards were failing to provide a good meaure. That's what they might have *thought*. But the important issue isn't what they thought, it is: what the reality was.

Observe the success of India-trained doctors abroad and here. In particular, realize that none quivers in fear while going under a surgeon's scalpel in any Indian hospital solely out of the apprehension that since there was no JEE for medicine, incompetents might have crowded into medical colleges. As to the effectiveness of the Boards at the higher end, observe further that until mid 1990s, medicine cutoffs used to be higher than the engg ones. So, even if we discount all the Boards-selected engineers, there still is a lot left staring at us in the face.

These IIT professors and JEE coaching class folks, together, have so badly twisted the landscape of available opinion that even brilliant and independent-thinking people have to be reminded of things like that---they can't get those on their own.

2. Re. R. Mashelkar and S. Patankar. Good catch. I should have thought of more recent examples. Can you think of any? Or do you mean to say there aren't any?

3. Mukesh Ambani worked as an applied (projects) chemical engineer during his early career. He took a long time to turn into a businessman---certainly much much longer than what so many JPBTIs incubating in IIT campuses do. UDCT had no donations-based admissions. Its cut-off would be highest in Maharashtra (because they had only a single and small program, chemical), closely followed by COEP (5 programs and biggest college), with VJTI trailing some way behind, then SPCE & Walchandnagar, finally by VRCE (now VNIT). ... People can be rich and also intelligent.

4. Venky attended 2 years after the matric (I guess) in your city, Baroda. I know that he seriously attempted JEE but could not crack it. I don't know if he also seriously attempted, at the same time, the medical side. ... Don't you think that in the 50 years of chest-beating, the JEE should have been able to pick out that one Nobel laureate we managed to school? Could the fault possibly be with the JEE? Are you *willing* to entertain that possibility? *able*? for how many nano-seconds?

5. Hmmm... Here, this is what I read: If IITs fail to train the _supposedly_ _initially_ good students sufficiently well, it's _never_ a failure on the part of the individual students. Ahem... That's a bit too much for me to take in.

If the cohort is so bent on going to the USA, and if it is so good on native abilities, then the JEE-GRE correlation should be great, right? Or do you mean to say that _native_ abilities significantly erode once the student enters an IIT? What about correlation in our times when every JPBTI was so keen on going to USA? (If in doubt, ask Kanwal Rekhi/Vinod Khosla/Virginia senators).

6. An example: You asked for an example. I could give one, but only from my times. I don't know the extent to which _today_'s Board exams resemble those of our times. To quantify that extent, I would need raw data. Historical as well as today's. Not just exam papers but also student performance data (the distribution). Only then could I be confident about the extent to which the examples I could cite, would be relevant today.

@Sriram:
Thanks for your honest feedback and good wishes. I sincerely appreciate both.

@Prof. Sanghi:
Thanks for running the third part of my previous comment even if belatedly.


All:
I will be traveling without internet connection for a few days. Will check back after returning, but, guess, all my points are already over, either at my blog or here or at the old LinkedIn thread. (In fact, many points got repeated many times over!) So, though I will make sure to check back, I don't expect to write further here once I return from that short tour.

Bye and take care,

Ajit
[E&OE]

aaa said...

another issue which needs to be brought into public domain is the manner of normalisation of marks across various state boards, cbse, icse etc. It was earlier reported that to break the deadlock that is to differentiate between two students, percentile will be taken upto 9 decimal points. To a layman, it means that a student with percentile say 98.000000002 will be graded higher than another student with say percentile of 98.000000001. Look how much difference even 1 mark in class 12 standard will make in the life of a student.(with all its follies of cheating, favouratism, subjectivity in marking, leakage of question paper s etc) This method has been approved by a Minister who has done away with marks in class X and brought in the grades to reduce stress. (CCE system is good or bad is another subject altogether). I hope there are sane voices in this world who will resist this impractical method

aaa said...

another issue which needs to be brought into public domain is the manner of normalisation of marks across various state boards, cbse, icse etc. It was earlier reported that to break the deadlock that is to differentiate between two students, percentile will be taken upto 9 decimal points. To a layman, it means that a student with percentile say 98.000000002 will be graded higher than another student with say percentile of 98.000000001. Look how much difference even 1 mark in class 12 standard will make in the life of a student.(with all its follies of cheating, favouratism, subjectivity in marking, leakage of question paper s etc) This method has been approved by a Minister who has done away with marks in class X and brought in the grades to reduce stress. (CCE system is good or bad is another subject altogether). I hope there are sane voices in this world who will resist this impractical method

Ritesh said...

1.Jee may not be flawless but it does't mean that bring a new system that is worse than the present.
Many people comment that in US they've single paper so we should also adopt it but they don't bother to see that their school system is much better than we have and above all it is fair and free from corruption.

2. Someone should remind Mr Kapil Sibbal that none of the students in this country gives 30-35 entrance exams.

aaa said...

well, some of the analysts in favour of proposed system have been bringing out that a student in the existing system has to appear for a number of exams.(30-35 No which is a false statement. a student may be appearing for a maximum of 5-6 selected exams) Well, then divide these into three categories, JEE, AIEEE and third one for state colleges as well as private colleges. At least this way, a student will have three chances in a year. Students who do not want to go for IITs will still have two chances in a year. Putting all eggs in one basket on a single day is calling for disaster for students who will be under tremendous pressure to do well. And by asking to give weightage to class 12 marks, the equation is changed in favour of Gunda Raj.
And if nothing is acceptable, then please ask IITs to conduct JEE based on basic level questions with say about 30-40% questions based on application of concepts so that every student has a chance without coaching. Enhance the cutoff for appearing in JEE to say 70-75%, if we want students to do well in board exams as well. Will, it not solve the problem? But for God's sake, donot give weightage to class 12 marks in any manner.
Out of 45 lac students of class 12, Mr Sibal's proposal may forcibly incline 5 lac students to school curriculum who are interested in JEE, what about other 40 lac students. How will Sibalji will make them study in Schools where there are no facilities, teachers etc.
Why not also consider giving weightage of class 12 marks to Medical Entrance, MBA Entrance, CA, CS, BSc, MSc etc, Does it not matter for these courses?

Saswata said...
This comment has been removed by the author.
aaa said...

Oh i forgot to add, Why not give weightage to 12th marks for the most prestigious exams IAS, IPS, IFS

And what about considering marks of class 12 for entry into politics?

I think the last one will make a good difference to the polity of the country?

An Indian said...

It is all over media now ... she has shown that she has the guts and she can get things done ... can somebody make a representation before her? ... newspaper report says that she is willing to listen ... this is to prevent backlash to academic community which somebody has pointed out ... happy to see the brave children Mother India would be proud of ... who have shown tremendous courage to confront the high and mighty ... it is patriotism redefined ... charcters do not date back to bygone era ... youth of India now holds IIT faculty as their hero ... salutation to the brave souls ... for you the world is still habitable ... more so to Saswat Sir of IITG, hat's off!

Piyush said...
This comment has been removed by the author.
Piyush said...
This comment has been removed by the author.
Piyush said...

ajitjadhav:
"Don't you think that in the 50 years of chest-beating, the JEE should have been able to pick out that one Nobel laureate we managed to school? Could the fault possibly be with the JEE? Are you *willing* to entertain that possibility? *able*? for how many nano-seconds?"


I don't see why you would expect anyone to not be willing to entertain the possibility that the JEE is not perfect. The owner of this blog seems to have done this often enough in the past: http://dsanghi.blogspot.com/2010/09/designing-objective-type-test-like.html

Also, if I am not wrong, Prof Sanghi is certainly of the view that JEE is not perfect, at least in his public writings. The argument here is not whether the old JEE was perfect. The argument here is that the proposed changes do nothing to address the imperfections, while introducing new ones (unfairness/arbitrariness in the scaling procedure, for example).

Just to take one instance, a common refrain is that the proposed changes will remove reliance on coaching. Not so long ago, I was a high school student in UP, and a majority of students from my school (which was, incidentally, quite good: the Physics teacher knew her physics, for example, and was always trying to go the extra mile to get students to go and study basic physics rather than engineering when they went to college) took coaching even for the board exams, with some putting in more coaching sessions for AIEEE or JEE. At least in the city I lived, this was more or less the norm in other schools too. To me at least, this means that unless solid data proving otherwise is presented, the claim that the influence of coaching institutes would reduce when more value is attached to board marks is one devoid of contact with ground realities.

boris said...

It is always sad to see that most people always pitch in the "Nobel" when talking of the IIT's not realizing that the IIT's were made to produce Engineers to aid in the development of the country,pure science research for which the Nobel is awarded was not an initial agenda.

IIT's have started taking research seriously in the last decade and in Engineering I would be lying if I said they haven't done anything.Examples:

-Karmarkar's Algorithm

-Google's New Page Rank Algorithm

-Madhusudhan's(an IIT-D CS passout) Nevanlinna Prize

-Subhash Khot's Waterman Award.

-SUN Microsystems(Vinod Khosla)

-Rajeev Motwani's Godel Prize.

So atleast in CS where the best go to IIT alumni have proved that they have achieved.

Now coming to the fact that most medicos aren't taken via entrance well that was a long long time ago,in most states board marks and state PMT/CET marks are combined for admission so yes there is an entrance and not to forget AIIMS has a highly competitive entrance where in out of 80000 only 50 make it,thus making it even more selective than the JEE.

Contd...

boris said...

Contd....

Many critics go ga-ga that lets replicate XYZ countrys model for India,well every country has a unique system for eg:in India one can join Medical School right after schooling but in the US you need to have a degree and clear the MCAT entrance to go to Med. School.Russians have their model,a highly developed country like South Korea has a hellish entrance system.We too made our own model.

Like I said in my earlier post IIT's want the best the country has to offer hence the standard has to be high now when you take something like the boards which are a sprint where irregular preparation may lead to victory you are lowering the standards and when that happens people who should not have made it thanks to that they make it.

I have always noticed that whatever be the field when the question is to create a group of highly skilled people the selection procedure is always harsh be it Education,Sport,Elite Military units etc.

Take the International Science Olympiads for eg they want to identify students with exceptional abilities in Science and then see the problems of the IMO,IPho,ICho etc they are way harder than JEE level,let alone high school.Problems are posed that even some professors are unable to solve them.

The JEE was no doubt hard in the Mains era from 2000-2005 but then again its predecessors like JEE 1987,1995 were scary.The objective though criticized was beautiful in it's own way the assertion-reason,matrix match and more than one answer question tested concepts,analytical ability very well.Though some knowledge beyond high school is required but isn't that a quality of the above average student who wants to learn more than what is there to offer.

Saswata said...
This comment has been removed by the author.
Shishir said...

@aaa, Very interesting point. Seriously.Why don't they (MHRD) suggest to give weightage to Class Xth, XIIth & graduation marks for Civil Services Exams? This will also give a big , a really big incentive for school children do well in board University exams. In fact , it will serve additional purpose of rectifying University education (sans professional one) which no one takes seriously !

This will save IIT and at the same time strengthen boards also.And will also boost University standard.

In fact , this idea needs to be pursued in a public domain.

aaa said...

@thanks for liking the idea.

I hope it matters to people who really matter in this matter. !!!!

aaa said...

there was supposed to be meeting of IIT alumni and others with the PM at about 4PM as per reports. Has it happened amidst all the drama going on for Presidential polls?

rahul said...
This comment has been removed by the author.
Dheeraj Sanghi said...

@manu, I feel hurt. Is there anyone who knows of any case where the comment has not been approved. I had written in this blog that I am traveling, and hence there can be delays. Most of the time I am looking at things on a 4-inch screen, and when there are lots of emails telling me that there are comments, an email or two can get missed, and whenever I have the Internet on the laptop, I check back on the site, and approve all comments there.

In fact, even this has happened that when trying to touch "Publish" on my touch-screen, the "delete" word gets touched. And I have taken pains to locate the user (I don't get to see your email address), and requested him to type the comments again.

Please participate in the discussions with a positive attitude.

rahul said...
This comment has been removed by the author.
rahul said...
This comment has been removed by the author.
Anonymous said...

@Manu, Why are we trying to pull so many reigns together? I think this discussion has lost the track. The point is not what IITs are good or bad at? The point is very simple- Whether the process of undergraduate admission to IITs as suggested by ministry is better than classic JEE or not? Does it tackle the issue of coaching and stress-reduction more effectively than previous exams? Lets discuss logically and not try to pull everything else here. I request Prof. Sanghi not to publish comments which are off the line or try to attackk somebody personally.

Sriram said...

@manu

---
Anyway the point I'm trying to make is that the IITs cannot cater to the large and vast Indian Population ...
---

Do you realize that you haven't made any point that is even remotely relevant to the topic under discussion? The blog owner is the host and we are guests. The blog owner is fully within his rights to delete off-topic posts or ones that don't contribute to the discussion in a constructive way. There is no absolute right to post on someone else's blog.

The point of a blog comments section is to talk to each other and not past each other or make comments that are tangential or personal.

gautam barua said...

I have put up some documents giving information on the IIT JEE reforms, including my article in a newspaper defending the changes, in a blog I have created. This is a info only blog (no comments allowed) and you can reach it by clicking on my name in this post. I request Dheeraj's indulgence in advertising my inputs through his blog. I will be adding a document on the autonomy issue later today.

Piyush said...

Dr Barua, in your article, you say:

"Hopefully, in the medium and long term, all coaching will take place at one place, in the school where a student is enrolled. Whether coaching institutes tie up with schools or they themselves get licenses to teach and become schools, does not matter. Schools will be where all the learning will take place."


As far as I can see, this is the only place in the article where you address the question of coaching institutes (correct me if I am wrong). Sadly, your argument here is merely by assertion.


We have all already heard horror stories about dummy schools in Kota, and if I remember correctly, these served as the proverbial straw that broke the camel's back and called for a review of the effect of JEE coachings on school education. Why do you imagine that this situation would change at all when Board marks are included in the evaluation too? I presume the "Board marks" factored into the calculation will be those for PCM and one language (either the student's regional language, Hindi or English). Thus, the coachings, in order to maintain their hitherto profitable "dummy school" model, will just have to add a separate "class" for English. Unless the board examination itself includes some kind of a multi-point evaluation in a meaningful way (not in the current "everybody-gets-full-on-practicals" way), these dreams, which you acknowledge are long-term hopes at best, that schools will somehow gain ascendency seem destined to remain just that: dreams.

So why not work first on making sure the multi-point evaluations in board exam are meaningful first, so that that there is som hope that board exam really do measure a student's year-long performance at school? Or is the point just that you now want two year-end evaluations, instead of just the one JEE? Even though the other alternative is at least as much, if not more, susceptible to influence by coachings?

Unknown said...

Post 1/2

The rationales behind the reforms and how the implementation manages to evade the actual problem at hand can be summarized below:

1. More emphasis on school education: The point, I guess, is that we need the quality of schools in India to improve as a whole - and not just "engineering school" education, which sadly seems to be the emphasis here. The engineering-oriented students form only a fraction of the total number of students in higher secondary (I recall having seen 0.25 somewhere). Thus, if the fuss is mostly about improving the schools as the directors were claiming in the meeting I referred to in my previous post, and if the scheme works out exactly as intended, even then only 25% of the school system shall be catered to and not the whole of it.

Also, the resolution being proposed has less to do with attacking the real problems - the lack of infrastructure, the unavailability of good teachers, the promotion of rote-learning in exams, making the examination system corruption-free, uniform and fair etc. Rather, it is hoped that tweaking the entrance exam processes might have some magical effect so as to get rid of the aforementioned problems (a friend of mine compared it to broadening the gates of a train to solve the problem of overcrowding on the platform). If it is hoped that the students ought to take the boards seriously, we are bound to improve the board brand first. Otherwise, the situation is tantamount to the one, whereby the government forces the people to buy a brand in order to "improve" it, rather than improving the brand so that people go on to buy it.

2. Reducing stress on students: One exam, as Prof. Sanghi himself has argued will probably only increase the stress on students. The real stress is the stress of preparation. Multiple entrance exams perform precisely the job of reducing the pressure by allowing the students to have options at their disposal and not shift the pressure to one-day, one-exam. Imagine choosing a cricket team based on just one day's performance!

3. Normalization: In a country with 42 boards, different board syllabi, with subjective exams and a large number of teachers grading the answer scripts and where perhaps the marks awarded might even depend on the gender, the mood, the food etc. of the examiner, is perhaps not a very good option to rush to. The argument of statistics, one should realize does ignore some information, and this is perhaps where normalization could be unfair to a small percentage of students and not surprisingly the absolute number could be too large to ignore. If the absolute number exceeds even hundred, then it would be a failure of the entire process of admission. Also, the idea that board exams be corruption-free and uniform can't be dismissed or ensured by any JEE policy.

contd. in the second post..

Unknown said...

Post 2/2

4. Reduce coaching: Whether coaching classes should exist, whether they are doing a fair job or producing "machines" in the garb of students which sooner or later undergo deterioration, whether they should earn a license first or whether they should be earning the amount that they do is a different issue altogether. As a matter of fact, coaching is neither illegal nor unethical and also no what matter what the exam, if there are a lot of students dying to clear it, coaching institutions will flourish. Their existence or subsequent thriving is more or less a consequence of the society's wish to pay them exorbitant amounts in return for training that they are otherwise not able to get in schools. And, believe it or not, they have geared themselves up for the next exam, whatever it be JEE or ISEET! Plus, I can assure that the number of people going for coaching of boards far exceeds the number going for JEE coaching. If we are against coaching, we should be against all forms of coaching (coaching for boards, the medical coaching, coaching for IIM, coaching for UPSC etc.) and not just coaching for JEE simply because these people make more money.

P.S. - And those of you who have grudges with the JEE because you could not clear it, its indeed a pretty cheap reason to criticize it.

ammu said...

Dear Prof Barua,
I am an alumni of IITK and has been engaged with different schools of Gurgaon for training teachers. So far I havent found a single teacher capable of teaching science to bright minds of this city(90% students of any pvt gurgaon school will score 110+ in IQ test)(All these schools charge more than 10k per month including all expenses). The problem with the current board examination is that they cant have good question paper because the teachers currently employed in school wont be able to solve basic science problems themselves. Moreover, after years of interaction I am sure 90 percent of them cant even be trained. After weeks of classes on Reflection(Teacher development programme TDP as they used to call them) a class XII teacher with 4-5 degrees on her resume tells class VIII students that Law of reflection is not followed incase of regular reflection. These are not "anecdotal", they are common in almost every schools of Gurgaon. I have interviewed many teachers and "90 percent failed to answer answer number of moles of people in India". Teaching at schools are least preferred job, those who failed everywhere else tries teaching at school. An organization is known by its people, if you want to improve school education, then glamourise teaching and make it more paying.
Teachers of DPS of one metro are paid 18k approx but they are forced to sign 30k slip. Why do you people are agianst coaching, Infact India should be grateful to coaching they have improved the knowledge content(not education though). If I Have the money then I demand to be taught by best teachers, thats my right. There are many issues with teaching at school, please spend few months in school you will get to know why your proposed idea will worsen the situation. Initial few days I spent in staff room of different schools changed everything for me, it will do more to you since it seems you are a larger fan of schools than I was.

gautam barua said...

Guys, please do not lecture me on how schools are and how teachers are. I am in the management Board of a Kendriya Vidyalaya on campus and I have seen many "vernacular" schools in small towns, in cities.
The argument that the JEE proposal will not solve the major problems of schools is fine. Who is saying it will? Not us.
Including Board results is a SMALL, but hopefully very significant step towards giving more importance to school results and so to schools. IITs, being the leading brand in education in India owe it to society to do its little bit here. And it is a very small bit. As I have tried to argue so many times, it is not going to make much difference to the meritorious candidates.
The IIT system and those who have been in it and those well equipped to enter it are by and large a privileged lot and many of the arguments I have read are selfish. We are doing fine, the outside is terrible, why are you bringing in all the dirt into our clean environment where excellence is the only mantra? Why fix it when it is working fine? Working fine for whom?
The point about coaching and schools is that if coaching institutes teach all the school subjects (after all the exams are going to have basically PCM, language, and an elective (which could be one or two in a coaching institute), then a student will get all his inputs from one place only, and that is the coaching institute. There will be no need for him to go to any other place or enroll anywhere else if the coaching institute takes the very small step of registering itself as a school. So Coaching institutes will become another set of expensive schools! Coaching will then "stop"!

aaa said...

gautam sir,
it seems that you are defending the proposed format for no valid reasons. It may be that you feel it from the heart that this proposed system will do a lot of good. Unfortunatley, none of your arguments are solid enough. And an eminent person like you should support your hearty feelings with logic. Well in your final comments,what I can gather is that you think that in the long term all coaching classes will turn into schools and thus coaching will stop. Wow what a logic sirji.I hope and pray that people who think on these lines are very few and unfortunately, authority to adversely affect the lives of lacs of students, parents is in their hands. God save all of us.

Anonymous said...

GB Sir,
Please allow anonymous (or OpenID) postings on your blog.
Ex IITG Alumna

ammu said...

Dear Sir, Without any iota of doubt KV are one of the best schools. And they are absolutely doing fine, but how many of us are fortunate enough to secure a seat for our wards at KV. Regarding many good vernacular schools you have visited, I am sorry to say you need to visit more schools.

"The point about coaching and schools is that if coaching institutes teach all the school subjects (after all the exams are going to have basically PCM, language, and an elective (which could be one or two in a coaching institute), then a student will get all his inputs from one place only, and that is the coaching institute. There will be no need for him to go to any other place or enroll anywhere else if the coaching institute takes the very small step of registering itself as a school. So Coaching institutes will become another set of expensive schools! Coaching will then "stop"!"


According to you Registering a school is a small step. "IT IS NOT". You cannot set up a school only for classes XI and XII, For first few years one has to run schools upto 5th standard then upgraded to class VIII, you need to have minimum infrastructure setting:bills for setting good schools equipped with all essential facilities runs into crores. Moreover you have to employ teachers having BEd. MEd, degrees(A B.Tech Degree holder cannot teach in schools. For teaching XI and XII one has to be minimum PG and BEd), most of the fresh Bed and Med are unsiutable for teaching, those who are somewhat ok joins govt schools where they dont take classes. They send some proxy, yougnster barely graduate to teach in place of them. Private school are always short of good teachers. In a walk in interview 500 applicants appeared to fill 40 vacancies, despite lowering our criteria to lowest possible we could only select 15 candidates, out of them four left within next year once they were selected for Govt Schools. There is huge dearth of good teachers. Good people will join teaching only if they are able to earn "significantly" more than what they could from elsewhere. I say significantly more becuase teachers nowadays are not much respected, hence pay should compensate for the lost respect. An Infosys employee is more respected than a teacher in small town and cities.(Atleast this is the case in my neighborhood). If you succeed in your plan then whatever good teachers we are left with in coachings will also vanish. Then you will have to teach MTH101 to kids who don't even know value of tan(45) and flaunts 90 percent in Boards(its true). The basic difference between schools and coachings is, "In school management or the owner makes profit and in coaching the teachers earns profit". Apart from that I see very few differences.
By forcing coaching to register themselves as school are you trying to force 25% RTE reservation scheme? Most of the good coaching already have scholarship schemes. There is no need for imposing RTE on Coaching via covert means.
The purpose of IITJEE is to select the best brain irrespective of whether they come from moon or Kota.

Shishir said...

@Gautam, I read your 'Correct Answer'.
I am somewhat surprised at your following comments:
'..First of all, the reason that the rule for IITs is different is because of the opposition the original proposal faced from a section of IIT faculty and IIT alumni. The distrust shown by many of them of the School Boards system has no firm basis, and the objections have been mainly anecdotal...'
Yes, an institute of national repute has to treated differently , taken special care of just to maintain its special status.
Further, the great leaders of yore , political and academic , saw the need of keeping IIT (and other institutes of national repute)on a high pedestal and therefore granted reasonable degree of autonomy and IIT delivered. Creative minds, which IIT were set up for, need freedom. These minds need to be allowed to experiment and form their own rules. (And these minds serve nation beautifully in their way,which is equally important)

Prof Dahl and Prof Kelkar were given an absolute free hand to nurture IIT/K in its initial days.And nurture they did!

You must be aware of reasons of fall of mighty institutions like Allahabad University and HBTI,Kanpur.It was because they were made to function like an extension of the state govt's department.
Autonomy is not elitist, it is a plain and simple functional requirement of keeping an educational institution creatively evolving.
I'd like universities who have some minimum threshold of excellence to be made autonomous (yes, let me add here financial help doesn't necessarily entail academic control) so that they also get into a virtuous cycle.
Moreover, if you are so convinced about the efficacy of stratagem of improving school education by giving board exams' weightage to any popular competitive exam, and you have a courage of conviction,will you please stand up and say that this should be implemented for all UPSC and state PSC exams also?
I'm not so sure..

Unknown said...

How coaching centers are prepared for the new JEE check the following link(18 schools)

http://www.raoiit.com/

thanks to kapil sibal,now more schools are into this multicrore business $$$

Anonymous said...

@barua
Aha! you say now that even elimination of coaching is not going to happen and they r going to be more vital! can you explain finally what is the 2 measurable goal of this new reform?

Anonymous said...

@Gautam Sir

Sir,
You are a bit late to predict that coachings will turn into schools. It has already underway in Kota:

Bansal Classes -> Bansal Public School http://www.bansalpublicschool.com/

Resonance -> Disha Delphi Public School http://www.ddps.ac.in/

Career Point -> Global Public School http://www.globalpublicschool.com/

Rao IIT -> Annapurna Public School

gautam barua said...

@shishir. Of course IITs must retain their autonomy. I will fight tooth and nail to preserve the autonomy we have. But please do not equate the admission process of UG students to the general question of IIT autonomy. Fortunately for the IITs, the Act and the Statutes, give a lot of autonomy to IITs. I have personally fought attempts to dilute the Act and the Statutes in the late '90s and early '00s. Fortunately for us, today there is a coalition govt., and they needed 1.5 years to create new IITs, so amending the Act is out for the moment. For me, the admission process for UGs is a small matter. The best students will flock to the IITs as long as we maintain our standards. In any case, as I have shown in the documents I have posted, the change is not eroding any autonomy. It is in tune with the Act.
This is Dheeraj's blog. So please do not make this into a one-to-one with me. Unfortunately I do not have enough time to allow comments in the blog I have set up, and I do not wish to upstage this blog to which I am so grateful for allowing me to have my say.
@neha - thanks for the info; At least the deception of enrolling in a school in Haryana will go now. At they will teach English! Our students will be able to write full sentences now! Anyone heard of "Junior Colleges"? They exist in Maharashtra, Assam and possibly other States: schools for classes XI and XII.

Piyush said...

Dr Barua:
"The IIT system and those who have been in it and those well equipped to enter it are by and large a privileged lot and many of the arguments I have read are selfish. We are doing fine, the outside is terrible, why are you bringing in all the dirt into our clean environment where excellence is the only mantra? Why fix it when it is working fine? Working fine for whom?"

Another argument by assertion. I'll ask to you quote one such argument. All arguments so far have been that the the current changes will not not improve JEE, not improve board examinations, not reduce stress, and not change the influence of coachings. The question then is what on earth is it trying to achieve? You have to have a rather jaundiced eye to say that those advancing these arguments are doing this for "selfish" reasons.

Also, we already have a pool of "elite" colleges in DU that take admissions based on board exams. We have seen what ridiculousness goes on there (100% cutoffs!). Even the minister found some consequences pf their admission policies absurd. Has school education in the Delhi (apart from that in rich and elite schools) "improved" because of this rat-race in board marks? In what way?

Piyush said...

Dr Barua wrote:
"The point about coaching and schools is that if coaching institutes teach all the school subjects (after all the exams are going to have basically PCM, language, and an elective (which could be one or two in a coaching institute), then a student will get all his inputs from one place only, and that is the coaching institute. There will be no need for him to go to any other place or enroll anywhere else if the coaching institute takes the very small step of registering itself as a school. So Coaching institutes will become another set of expensive schools! Coaching will then "stop"!"

The keyword here appears only in the last line: it is "expensive". So much from someone who was claiming those opposing his arguments to be "privileged" and "selfish". How one can argue that having good education in a few more expensive schools and a few elite Central Govt schools like KVs (which is what we already have) would improve the state of school education throughout the country is beyond me.

Piyush said...

Dr Barua said:
"The argument that the JEE proposal will not solve the major problems of schools is fine. Who is saying it will? Not us.
Including Board results is a SMALL, but hopefully very significant step towards giving more importance to school results and so to schools"

This is a rather weird case of putting the cart before the horse. On the one hand, you say (or at least imply) that the school system in the country, in general, is in a bad state. On the other hand, you concede that the proposal would not improve it.

And then you argue that what we owe to the school system is not to improve it, but to just give it more importance? What tortured logical arguments are needed to go from "we can't improve it" to "so let's just increase its importance"? If there was a rickety bridge on a river which was in a state where no repairs could be made, is your solution going to be to route more traffic through the bridge?

Piyush said...

It is also a bit sad to see the one sane and important suggestion that has been made to remove stress and make the JEE more getting washed out in the cacophony of other arguments. The idea is that assigning ranks does not make much statistical sense, and that we should depend upon them as less as possible, probably just by means of a cutoff. In other words, the JEE (and anythings else that has to be added to it: AIEEE or board exams) should be used to decide only who gets in, and not what branches they get. At least at a place like IIT Kanpur, where everybody does the same courses (modulo on 0 credit course) this wouldn't make a difference at all. Most people should be asked to choose their branches after the first year (and here departments may introduce their own weights on courses to make rankings: CS might put more weight on Intro programming and Math courses, EE more on electronics, Math and Physics courses, and so on). Some students who are still not sure should be allowed to scour for possibilities for another year (where they might have slightly diminished possibilities, but that would be of their own choosing).

At least within the course structure of an IIT where almost everybody does the same courses in the first year anyway, this would not be so hard to put into practice. The rat race or higher ranks in the original JEE would also reduce since they wouldn't come into the picture anyway (or would, at lest, not come into it in a big way).

I believe such ideas have been in the air or a long time (perhaps even on this blog, and I apologize for "stealing" other people's ideas if that has inadvertently been the case). I can only wonder what arguments have been raised against them, and why no one has been advocating for these real changes.

Unknown said...

@Gautam: WRT first post on your blog.

The reasons that you have stated in your favour are precisely the ones which work against you.

1) how can ranks across mostboards be equated?
2) how is JEE entrance policy going to improve school education? (engineering-studies is only a small aspect of boards - there is much more to boards than engineering alone. Plus the problem of infrastructure, poor teachers, corruption, non-uniform grading can't be solved this way.)
3) not everyone who writes the AIEEE writes the JEE. Is this not undue pressure on students when we ask everyone to write the JEE?

These are not rhetorical questions as you ended up presenting and also conveniently chose an answer of your choice. Trust me, these are, but rather deep problems that plague the proposed reforms and every single claim of this kind needs strong backing. And I am sorry, till now you have offered none at any point or any platform.

Additionally, I don't get the fuss about coaching for specific subjects. Is coaching bad because its good when it comes to expertise on subjects and the schools are not? Or because we envy them because of the money they are earning? Suppose the schools in India were all A+ and assume that the coaching were A10+. Would you have a problem with students running off to coaching institutions to get training from ten-fold better "coaches"? It would be interesting to know your answer here.

Plus FYI, I can assure you that the number of people going for coaching of boards far exceeds the number going for JEE coaching. If the school+board system is such that people have to resort to coaching for boards, then how on earth is it being imagined that board+JEE/ISEET will slowly diminish the importance of coaching. Also, there is nothing unethical or illegal about coaching, and it would be very undemocratic if the "people in power" want to eliminate coaching entirely because it pleases their fancy to do so! It is certainly even weirder, in the light of the fact that they are doing a better job than institutions recognised by these very people in power to be centres of learning.

"Schools will be where all the learning takes place." - I don't really appreciate this. People don't learn just in a place but from their experiences, from their elders/parents/relatives, from friends, from accomplished individuals and from their idols. If you mean to say that all these should come and accumulate in their schools to teach them and that everyone must acquire a license before he/she can pass on his skills, it would only be ridiculous.

Looking forward to an earnest answer from you, and kindly avoid distractions/diversions,

Vivek Lohani,
Alumnus IITK

Piyush said...

From Dr Barua's information on the website:

4 IITs (Delhi, Bombay, Kanpur and Roorkee) are against using board marks for the final ranking. Among thse, only Roorkee has considered using gibing a weight to the board marks at all. Madras's preferred option also is to use board marks only for screening. Kharagpur in unclear.

Yet, if I understand Dr Barua's claims, it is that the 'majority' opinion among IITs is the solution agreed by IIT Guwahati? What field are we counting in?

Ankur Kulkarni said...

Prof Barua, you are basically proposing that the new IIT entrance be designed so that it serves as an instrument for improving the school system. I think this is a good idea on the whole, but it is unfair on the IITs. The IITs and the JEE are not responsible for the defunct school system, and it is not fair that the JEE be compromised in order to improve the school system. Why is it fair to the IITs and its faculty to be saddled with this burden, which is in principle to be borne by the HRD minister, not the IIT faculty?

Although this is not fair, the IITs can, if they feel so, show some largeness of heart to help out in this social engineering. I think you and the ministry need to recognize this and change your approach. You need to stop trying to make this new exam look like a cure for the JEE. You also need avoid claiming that you have a solid mathematical case for your idea, which I doubt you have. Instead you and the ministry should just make a request for cooperation to the IIT Senates. A request to incorporate XII marks in some way, for the greater good of the school system. And then leave it for the Senates to decide, thereby respecting their autonomy, their intelligence and their concern for the society. The current approach is arrogant and it reeks of a self-righteousness that cannot be justified, given the methodological flaws that your proposal has.

ammu said...

Dear ALL,
Could someone get the income profile of Parents of top 10 Delhi University parents(top 10 among themselves) and IITs.The data will be very useful.
Inclusion of board marks may favor the rich. I feel very few people from lower middle class get into good colleges of DU(data will convince us all). But IIT is place where students from all background can be found. This is certainly due to the levelling effect of JEE.

One more fact I would like to share.
Sanghi sir please take note of this point.

In Bengal Boards, the students are allowed to write answers in Hindi, Urdu, Bengali and English upto class X standard. However for Class XII only Bengali and English are allowed. NAturally those who have appeared as hindi or urdu candidate for class X are at disadvantageous position than their bengali fellow in class XII. By what formula on earth parity can be ensured to all Bengal Board Students.

aaa said...

well, why not increase the scope to most stupid level? Why not take into consideration marks of class 5 onwards till class 12 for entry into any undergraduate course? will it not ensure that the child is put through hell and every one exploits him/her but so what, kapilji's dream of giving importance to school education will be fulfilled. And thereafter,why not equalise the marks of various colleges for various streams for admission into post graduate courses. Why not have one exam in class 5 across india for entry into class 6 and one central exam in class 6 for entry into class 7 and so on. And if same intelligent people who are in favour of proposed system of JEE object, ask them why not? It is surprising to see the logic or should we say illogic, by so called experts who are favouring the proposal ?

aaa said...

well, why not increase the scope to most stupid level? Why not take into consideration marks of class 5 onwards till class 12 for entry into any undergraduate course? will it not ensure that the child is put through hell and every one exploits him/her but so what, kapilji's dream of giving importance to school education will be fulfilled. And thereafter,why not equalise the marks of various colleges for various streams for admission into post graduate courses. Why not have one exam in class 5 across india for entry into class 6 and one central exam in class 6 for entry into class 7 and so on. And if same intelligent people who are in favour of proposed system of JEE object, ask them why not? It is surprising to see the logic or should we say illogic, by so called experts who are favouring the proposal ?

mkg said...

Prof. Barua,

I read your blog. I am disappointed to see this discussion is now reduced "legal arguments".

I am not a lawyer. I read selective ordinances of the IIT Act (on your blog). I read "33(2)(a)" as well as "33(2)(b)". Only "33(2)(a)" has clear mention of "admission standards". "33(2)(b)" applies to other things and in the end "other matters of common interest". In your response, you mention that "33(2)(b)" applies to "matters of common interest". You forgot to include the word "other".

So, the way I understand it (and, again, I am not a laywer), that "33(2)(b)" applies to those issues which are not covered in "33(2)(a)". Is that wrong?

I hope we limit the discussion based on facts, data, research and analysis and not let it degenerate into insults, innuendos, and question each others motivation.

Regards.

Unknown said...

i have a little insignificant doubt?

say we have a gov in UP who imposed ban on mass copying and result is 17% only and next year we have another govt who declared(cheaters will not be jailed ,and it happened too in actual), the result goes up to 88% with overall score and highest marks,normal distribution etc etc changes too much.
Now can anyone from IIT or ISI kolkata explain how to normalize this result fairly???

I find this little doubt nagging my brain too much.. so better put and ask prof Gautam and Prof barua to soothen my worrying mind.

Thanks,

Gaurav

Unknown said...

@Ankur Kulkarni: Indeed, you have hit the heart of the matter and done a fair job elucidating the diplomatic, political and academic intentions and how the former two camouflage the latter in order to cater to the ego of the ministry. Kudos!

What I essentially don't understand is why the IIT-council comprising of IIT-directors would be hell-bent on embracing such "suicidal" measures. I mean, come on directors, how much IQ does it take to see that most of the arguments that you are resorting to are utter nonsense!

I can understand it from the ministry's perspective though, because from there its only tantamount to imposing "murderous" reforms, and mind you not "suicidal". There is however no denying, whatever the perspective, that in the long run these are bound to be catastrophic.

Chandresh said...

Dr Barua is making some interesting points:

1) Inclusion of normalized percentiles sends a message to students to not ignore board exams and non PCM subjects.
2) Inclusion of normalized percentiles across boards will have only a marginal impact on the proposed composite ranks as the difference in percentiles will add a few delta differential “bonus” marks to students who do not ignore board exams.
3) The addition of a few “bonus” marks to students who do well in board exams is at most equivalent of giving a head start of 1-2 questions in the JEE exam. In a majority of cases the differential advantage will be in fractional marks.
4) The above differential head start to sincere school students will predominantly affect the ranks in the lower most quartile of JEE qualifiers.
I would think that these are interesting and fairly persuasive arguments.

The real issues seem tangential to the incorporation of percentile board mark in rankings. The actual problems seem to be:
1) Why not delay the reforms by a year. Is it because anti-reform groups will sabotage the reform? Or do elements of the establishment have a hidden agenda and is presenting a fait accompli? We have no conclusive evidence either way.
2) Is the switchover too hurried and affect the students appearing in the coming year? The evidence seems to be yes, it will negatively affect the student community at least in the coming year.
3) Does the new system carry more risks like dependency on external factors like all board results being fair & published in time? The risks do seem greater compared to present system that is fully controlled by the IITs.
4) Has AIEEE been more progressive in adopting modern exam management practices? Yes, IITs might need to learn innovativeness from AIEEE which administers a more heterogeneous and complex exam.
5) Will the reforms promote equity between different boards, centre and state systems or be unjust? Both sets of arguments have merits.
6) Will the reforms promote better opportunities to students from economically weaker sections, rather than the middle class? Difficult to predict whether the reforms will have an impact on poor vs middle class argument.
7) How to inculcate dreams and values in students?
8) How to push IITs to the next level of excellence?
5) How to push boards to the next level of efficiency?
9) Will societal problems of school education, coaching, plight of underprivileged, selecting the “intrinsically best” for IITs etc be solved by reforms. The evidence seems to be no, not significantly.
On balance, is it not better for the academic community to unite & pull back from the brink and come up with a solution based on scholarly evaluation? But this approach needs trust and integrity to carry out genuine reforms in the next academic year.

chitta said...

The folly of normalizing percentiles across various boards:

Dheeraj makes the point in http://dsanghi.blogspot.com/2012/06/jee-2013-why-discriminate-against.html but perhaps it needs but perhaps the following examples might might strike a chord more easily.

1. Consider noralizing the percentile of AIEEE and current IIT JEE by treating JEE top X% as equivalent to AIEEE topX%. It should be obvious to anyone that this is incorrect. Not just IIT JEE rank top X (say X =100) is no way equivalent to AIEEE rank top X (say X =100) but because say double the number of people take AIEEE than IIT JEE, the percentile approach will treat top 2X (200) of AIEEE with top X of IIT JEE.

2. Now consider normalizing the percentile of IIT JEE, with various state entrance exams by treating IIT JEE top X% (or top X percentile) as equivalent to say Tamil Nadu or West Bengal or ABCD state JEE top X% (or top X percentile). It should be obvious to any one that either is incorrect. This can be verified easily to be incorrect. Just check the JEE ranks (convert them to percentile) of the top XYZ% of any state JEE.

(Note that CBSE and ICSE have all of India as their scope -- similar to IIT JEE and AIEEE; while the state boards have their scope the individual states -- similar to the state JEEs.)

Unknown said...

@ajitjadhav: I believe @Piyush above has done a very decent job of dissecting Dr. Barua's arguments. Please do go through his comments.

In any case, I have collected my thoughts, and posted them as three articles on my blog: http://wirmuessenwissen.blogspot.in/

I request (and thank in advance) Prof. Sanghi for allowing me to direct users to my blog.

In any case, if anyone has any comments on my blog, please post them on this (i.e. Prof. Sanghi's) blog, so that a bigger audience can see them and reply to them.(@ajitjadhav, I have compiled an FAQ section in part 2/3, in which I have tried to answer some of your questions.)

Anonymous said...

@Tanmay Mudhokar June 17, 2012 8:44 PM

If you request me to go through your blog, I will. That's *my* word of honor. (BTW, you also have my email ID, don't you?)

Ajit
[E&OE]

Sriram said...

@Tanmay

Your blog posts are well articulated. One minor quibble: the JEE is not a direct test of "intelligence" (as an IQ test might be). It is test of knowledge. That it is also a proxy for intelligence is likely but that is indirect. A person with high IQ but no knowledge will flunk the JEE.

In the real world, verbal ability (written and spoken) is important. It would be nice to ensure that IIT students have adequate to excellent verbal skills when they graduate. Clearly, in your case, the written skills are well above average. But I have to wonder about the typical IIT graduate (likely, average or below average, relative to international norms for university graduates from reasonable places). All elite American universities demand that their undergraduates take a few 1-credit writing courses (even students who have maxed out the SAT verbal). Writing is a skill that needs practice and feedback and the marginal effect of a few writing courses can be out of proportion to the time spent.

Unknown said...

@Sriram: Two minor quibbles with you:

1) There is no exam that tests only and only IQ, and likewise there is none that tests only and only knowledge/information stored in the head.

The idea of an IQ test itself is rather vague and if you strongly feel that it is indeed a measure of intelligence it would be interesting for you to know Feynman's take, who we all know was intelligent, but had a low IQ!

2) I believe that an exam like JEE ought to test the preparation, the willingness and the motivation to pursue science and engineering. As such it is immaterial if the person had a high IQ but didn't work hard. As one of my professors simply puts it, "It certainly doesn't matter to me if there were people who were smarter than Einstein but not passionate enough, because he was the one who actually solved the problem of relativity and not these guys."

Unknown said...

@Sriram: Thanks for your input. I am sorry if I have created an impression that JEE tests a student's IQ. I assure you, I do not believe that is true, and what is more, I do not believe that IQ is what we should be testing. I shall not get into the nature-versus-nurture debate here, but let me merely point out that innate ability, by itself, is not sufficient to succeed in any field whatsoever. Therefore, we should be testing for a combination of hard work and innate ability (which is what we already test in JEE). The relevant question is: How close is this tested combination to the combination required for success? This is a fairly valid and open question.

As to your other point about IITians lacking soft skills, I agree completely. I would even go so far as to suggest that the JEE should include a one-hour essay after the Mains, and that performance in this section could be used as a cut-off, not for ranking.

There could be measures adopted within the IIT system as well. IITK conducts an English test for all students at the time of admission. If a student is found deficient, he has to do a course on Basic Written and Spoken English in the place of a regular Humanities course. Some form of participation in debating societies could be made compulsory... the point is that teaching English to inarticulate students is far easier than teaching regular subjects to students who have never seen anything other than State-approved textbooks. (And if 'State-approved textbooks' seems like an Orwellian phrase, it is deliberate.)

aaa said...

well, I hope that sensible decision to immediately announce postponement of proposed examiation pattern to Not earlier than 2014' will be announced soon. Politicians,bureaucrats, IIT directors, councils, senates, alumni, all need to understand the importance of clearing the anxiety in the minds of students preparing for IIT JEE for 2013. They can devote one or two days out of their busy life/schedule, and decide on this matter within next 2-3 days.

Unknown said...

@Gautam:
"The argument that the JEE proposal will not solve the major problems of schools is fine. Who is saying it will? Not us.
Including Board results is a SMALL, but hopefully very significant step towards giving more importance to school results and so to schools. IITs, being the leading brand in education in India owe it to society to do its little bit here. And it is a very small bit. As I have tried to argue so many times, it is not going to make much difference to the meritorious candidates."

Now, as Piyush has nicely explained above, this means if you can't improve something, just put more premium on it. Also, I am given to understand that the a step of this kind is "big" or "small" depending on its significance or insigficance. So, significant and insignificant in the same sentence about the same object!



"The point about coaching and schools is that if coaching institutes teach all the school subjects (after all the exams are going to have basically PCM, language, and an elective (which could be one or two in a coaching institute), then a student will get all his inputs from one place only, and that is the coaching institute. There will be no need for him to go to any other place or enroll anywhere else if the coaching institute takes the very small step of registering itself as a school. So Coaching institutes will become another set of expensive schools! Coaching will then "stop"! "

All I can gather from here is that the only problem with coaching is that they don't teach one language course and one more additional subject. Now suppose that it starts happening and mind you, "coaching" institutes can easily remain "coaching" institues teaching 5 subjects (and not turn into schools). Is the problem resolved? Given Mr. Barua's comments, he would probably say "YES". Academicians would, however, say a big "NO" because all of them would be concerned about the fate of school education and less about the existence of "coaching". And even if we turn a bit optimistic and hope that coaching institutes will eventually turn into expensive schools, have we any done any good to school education in the country? I fear that the answer here is going to be a bigger "NO".

The contradiction among the various stands assumed by Mr. Barua stems from the fact that in spirit he seems to be against "coaching for JEE" (and not simply "coaching") - the grounds mostly being that coaching institutes somehow "program" students with lesser innate talent to qualify the JEE (these views have been put forth at different platforms). Thus, on one hand he wants to exterminate them, but since that would be undemocratic, he wants them to turn into expensive schools! What a glaring contradiction that is, because we wanted them to go extinct for a reason and lo and behold, here they turn into schools!



"At they will teach English! Our students will be able to write full sentences now!"

And what makes you think that the schools will nail it here when they failed abysmally elsewhere?

Sriram said...

@vivek lohani

This is seriously off-topic but Feynman was a prankster and it is highly likely that he dumbed himself down on that school IQ test at age 15 and subsequently loved to tell that story. That's my guess. Steve Hsu has his own take on this issue that you may find interesting.

http://goo.gl/vBrPy

Yes, there is no one magic IQ test. But such measures tend to be positively correlated and the common (latent) factor is what is often referred to as "g" (fluid/general intelligence). Sadly, "g" peaks in the early 20s and there is a long, slow decline (I am many decades past my peak). Entrance exams, since the 1920s, are essentially IQ filters because threshold levels are needed to learn materials of varying complexity.

Sriram said...

@Tanmay

--
teaching English to inarticulate students is far easier than ...
--

Such considerations should drive curriculum development. Namely, the marginal effect of any input (on the output). A whole series of specialized courses may be needed to reach a certain level but often just one course (or even a short course lasting a week or two) on some topic can make a huge difference. Most curricula focus on the former and don't consider the latter.

rahul said...
This comment has been removed by the author.
rahul said...
This comment has been removed by the author.
Unknown said...

@manu: I'm afraid I disagree with you very strongly. The whole point of the JEE is that there can be no official textbook, there can only be an official (not necessarily strictly adhered to) syllabus. The point is to get rid of the Indian mentality to think in terms of fixed text-bookish questions and syllabi. No one should feel satisfied that they have 'covered', say, electromagnetism when they have gone through 's learned treatise on the subject. Giving a finite amount of official 'study material' will only be an extension of the Board mentality, which the JEE is supposed to filter out.

@sriram: Sir, I agree with you thoroughly. I think the IITs do need to pay more attention to this.

VaibhawC said...

What IIT K senate did was what we had always been craving. Well, when all this hurly-burly is going on, we got the right time to put JEE on track as many of us feel even the current JEE is not GOOD. Without the mains exams this JEE is nothing. Either we further put the interview thing or we take more JEEs in a year to "reduce pressure".. we still have to work on it and I wish that discussion over JEE continues further even after all this is over.
Of course, we have to here choose between importance and urgency, quality and feasibility.

VaibhawC said...

What IIT K senate did was what we had always been craving. Well, when all this hurly-burly is going on, we got the right time to put JEE on track as many of us feel even the current JEE is not GOOD. Without the mains exams this JEE is nothing. Either we further put the interview thing or we take more JEEs in a year to "reduce pressure".. we still have to work on it and I wish that discussion over JEE continues further even after all this is over.
Of course, we have to here choose between importance and urgency, quality and feasibility.

Piyush said...

A news story that might be of interest"
http://www.thehindu.com/education/college-and-university/article3543118.ece?homepage=true

What Dr Sanghi and others predicted about the movement across boards seems to be already happening.

Ajay said...

Congratulations !!!
Interesting news

http://www.ndtv.com/article/india/iit-entrance-exam-row-breakthrough-soon-233146

aaa said...

i hope whatever decision is taken is implementable and obviously nothing should be changed for 2013. I sincerely pray that people whom the God has given the authority to change lives of people will behave in a more sensible manner.

aaa said...

well I just watched news on NDTV hindi channel. While there seems to be some agreement on not considering marks of class 12 and having JEE mains and JEE advanced on different days, there is no clarity as at now as to when the new system will be applicable from. I have got this uncanny feeling that Mr Sibal will not give in on all the issues. HE will probably insist that it should be made applicable from 2013. Well, Mr Sibal and IIT Senates,Alumni will have something to cheer about but what about the students who are going to appear in 2013? Why should changes be brought about at such short notice for them? Will any one think about them?

rahul said...
This comment has been removed by the author.
aaa said...

Well, from the news articles it seems that a compromise formula will be worked out and the system will be made applicable from 2013.

What a shame. Both sides are probably reacting in a manner as if it is a question of bartering. No one seems to be understanding that changes are welcome if made in time and with due transparency and logic. It seems that students taking exam in 2013 will be the worst sufferer. Once some decision is taken, it will still take lot of time for the formats to be decided for JEE mains and JEE advanced.

Well, obviously the very obvious thing to do is to first announce that it will be applicable from 2014. But it seems a common man does not understand the complexities that a minister, bureaucrat, IIT directors, alumni, federations can peep through.

We as a nation are good at first complicating the things, then look for solutions and then further complicate the issues. Well done all.

Just to remind the well wishers of this proposal, it is the same batch of students who will be appearing in 2013 exams who were the first batch for CCE experiment started by Mr Sibal. Well, all changes are being brought about with one batch of students only. Any specific reasons.

And in the end, What is the big hurry to implement it from 2013. Any answers??????

Ajay said...

Sibal offers trade-off to IITs

http://www.telegraphindia.com/1120620/jsp/frontpage/story_15633511.jsp#.T-EhefIa570

Ajay said...
This comment has been removed by the author.
Ashok Yadav said...

A view from other side, a 16 year old student on this matter...
http://www.ndtv.com/article/india/what-iit-means-to-me-a-16-year-old-s-take-233663

Anonymous said...

who has bandwidth to conduct such a test?
http://www.indiamag.in/from-2011-aieee-to-go-online.html
is there any lobbying from testing agencies to begin in 2013 itself? god knows

aaa said...

mr sibalji has still not tried to understand the other's view point. If there were no issues, then why would PM meet IIT senate, faculty and alumni representatives? I am sure better solution would have been suggested by these learned people to PM and to Mr Sibal. Now to say that proposal will be implemented from 2013 and it is for the IITs to suggest better proposal is like being adamant and illogical in my humble opinion. First logical thing would be to announce the postponement of proposed changes with effect from 2014 and then have a transparent discussion on the subject.-But then who cares and more importantly, why would someone care for students specially those appearing in 2013. Do we not know that these students have been preparing for the existing format of exams since say their 9th or 10 th class. Will it not be unfair on them to bring in major changes when these studens are in their 12th. I wish and pray some one cares.

And lastly, IITs would give a better proposal, if Min of HRD will ask them. Has anyone asked them after their meeting with the PM?

aaa said...

well, it seems mr sibal is adamant on implementation of modified propsoal from 2013. So, the following may be implemented in my view:-
top 20% from each board to be eligibile for JEE (Main)

Top 10% from JEE (Main) to be eligible for JEE (Advance) as against 50,000 which in my view is too small.

JEE (Advance) to be conducted after declaration of result of JEE (Main)

JEE (Advance) should be MCQ as per existing format for 2013 and may be changed to subjective format for 2014 as desired by IITs.

what say ? It will meet both the requirement of ministry as well as IITs. And more importantly, will not disturb students presently in class 12 much.

Any takers?

Manish Verma said...

@aaa, As I understand, JEE-Main is the main exam for all other CFTIs initially and for States as and when they join and for private colleges as and when they join. Therefore, the exam itself may not have any strong requirement/weightage for 12th performance. It is the institutes admitting via JEE-Main that may have requirement/weightage for 12th. For example, NITs can give certain weightage of normalised 12th along with the weightage of performance in JEE-Main. Some other college can keep some requirement of normalised 12th in place of weightage.

As far as IITs are concerned, they will select students through JEE-Advance and they can have their own minimum requirement for 12th board since weightage in a bone of contention.

It is clear that the govt./Council wants the focus on 12th. Keeping suitable percentile as requirement does seem to achieve this as a student would like to be among high rankers in his/her board to come within the percentile bracket.

One can of course debate on how much percentile should be kept for IITs like 80 (top 20%) or otherwise.

Many among IIT faculty are talking about deterioration of quality of input after JEE was made completely objective, i.e. from 2006. While, I am unable to find data to support this, assuming their assertion is correct, it is time that JEE-Advance has substantial component of subjective testing as well.

Unfortunately, 2013 aspirants are facing more uncertainty. As long as the uncertainty remains, they need to console themselves considering the fact that it is the relative performance that matters and all aspirants are in the same boat.

aaa said...

@Manish

my view that why give weightage to 12th marks even for NITs. Let these be used only as a cut off even for JEE (Main). there is lot of subjectivity in marks for Class 12 as discussed at a number of blogs. therefore, even by enhancing the cutoff say top 20% only eligible for JEE Main, will we not achieve the desired purpose of making school education more relevant. as far as IITs are concerned, students should be required to come through two screenings,first school marks cutoff and then top 10% of JEE (Main). In 2013, JEE Advance should be MCQ only as it is too late to fiddle with it. but from 2014, JEE advance can be made a combination of MCQ as well as subjective or completely subjective.

and lastly, argument that everyone is in the same boat that is why even if the system is changed from 2013 should not be taken too far. i hope people understand the emotional stress on students in such a scenario. on the contrary, if everyone is in the same boat then why not do it from 2014 which everyone will readily agree to.

Arpan Maheshwari said...

Hopefully IITD and IITK would hold the exams together. Good news !

http://timesofindia.indiatimes.com/india/IIT-Delhi-rejects-Kapil-Sibals-common-entrance-test-proposal/articleshow/14322175.cms

aaa said...

mr sibal can still become a hero by announcing that changes will be implemented from 2014 and end the uncertainty in the minds of lacs of students and parents.Then discuss, changes in a transparent manner.

we Indians are very good at making simple issue an ego issue and behave like stupid people where we ought to make ego issues. For example, we should make it an ego issue that we will have better hygiene, sanitation in our public places, railways etc. We should make it an ego issue that school education will be improved in next say 5 years. But no,we will take a small issue to a level where it will be difficult to back down even if one wants to.

the art of life...!! said...

Sir,what is the current situation for the eligibility criteria for 2012 passes students and sir, is their any chances that this pattern will not consider for session 2013-14.