Friday, June 30, 2006

An analysis of the results of JEE-2006


The Ministry of Human Resource Development (MHRD) has issued a press release about the results of the new, improvedTM Joint Entrance Examination (JEE) conducted in April 2006 by the Indian Institutes of Technology (IITs). It's clear that the Ministry wants a positive spin on the outcome, but it's equally clear that a lot of the news is not all that great. Let's take a look:

Bias against first-timers:

The number of candidates qualified in their first attempt is 2,761, which is 43.50 per cent of the total qualified candidates. This proportion is significantly greater than the corresponding value (28.49 per cent) in JEE-2005. The high proportion of first timers reflects the success of the changed pattern of examination in JEE-2006 in getting more number of students who have been equally good performer in qualifying examination (10+2). Hence, the main purpose of introducing a new examination pattern emphasizing on the importance of school education has been successfully fulfilled. ...

Bias against rural students:

The proportion of successful candidates belonging to towns and villages has also increased to 30.67 per cent [1943 students were from towns (1328) and villages (615), while 4400 were from cities], as compared to 28.02 per cent in JEE-2005 while the percentage has decreased in case of cities from 71.98 per cent in JEE-2005 to 69.37 per cent in JEE-2006. The proportional increase in the percentage from smaller towns further emphasize the success of the new JEE system and indicate reduced dependence on coaching centers which the candidates from town and village have no access to.

Bias against girls:

The application fee for female candidates was half (Rs.300/-) of the fees for male candidates (Rs.600/-). In this respect the new JEE system has also ensured a higher participation of female candidates as evident from the total number of registered female candidates of 58,997 in JEE-2006 as compared to 29,291 in JEE-2005. However, there is only marginal increase in successful female candidates as compared to JEE-2005.

Clearly, there has been significant 'progress' only in the share of seats that went to the first-timers. Progress made by non-urban students is small, but it's at least in the right direction. In the girls' share of IIT seats, the progress may even be deemed 'negative', since their number increased only marginally in spite of a near-doubling in the number of girls who applied for (and presumably, took) the JEE.

The JEE still has a long way to go before it can be held up as an ideal for the other entrance exams to emulate; right now, I would view it only as a deeply flawed exam with in-built biases against girls, first-timers and the rural and the poor.

IMHO, IITs should strive to convert JEE into an exam that fulfills -- at the least -- two primary requirements: (a) It should be a standardizing exam, in the sense that it should allow one to compare the relative levels of different school board exams, (b) it should be a standardized exam, in the sense that its results are less noisy and more predictable. Better yet, the IITs should merge it with AIEEE conducted by CBSE, and seek to make that unified exam conform to these two requirements.

* * *

Let me list a few related posts for future reference:

How good are our entrance exams in 'discovering' merit?

Critiques of JEE by two professors of IIT-K

Women in IITs

Consistent performance

Entrance Exams

4 Comments:

  1. Sourav said...

    As a pass out from IIT Madras, I have been through and seen the JEE system closely.
    From my experience,I can say that the effectiveness of new system can be best understood in terms of the number of qualifying students who are not a part of the major coaching "factories".

    The rural urban divide is not a proper indicator in such cases. While it is a fad among people from semi rural areas in North India to study in the Bansal classes at Kota, or at FIITJEE in Delhi,people in metropolitan cities like Calcutta or Chennai form very small proportions of these coaching factories.
    Since for the last few years,IIT selections(read "merit")had become a monoploy of such "factories"(We used to joke about the "Kota Textiles" in our IIT hostels),it is important to know if the improved system is able to cope up with this problem.
    I understand the problems of the Government in furnishing such details pinpointing particular commercial organisations,but such statistics,if provided by any source,shall be better identifier of the efficacy of the system.

  2. Anonymous said...

    Hi Abi

    The increase in the first-timers is definitely good. From how much ever I could notice, it is usually the second- and third-times who come to campus burnt-out and lead a retired life through their four years. The broadband LAN in each hostel room, a cultural festival in the even sem and a technical festival in the odd sem make sure that some of them stay comfortably numb.

    The modified JEE hopes to reduce the proportion of coaching class students and have more of the 'original' ones who are not from major cities. JEE-2006 had a one time exception to let whole bunch (3 lacs as opposed to usual 2 lacs) to appear. Lets wait for next year's results before drawing strong conclusions.

    Cheers
    Phani

  3. yogesh upadhyaya said...

    Abi:

    Thanks for your excellent analysis of JEE, published time to time. I offer my views as follows:

    1)Everyone is bashing the successful exam, which may be the only qizz-type engineering entrance exam in the world. Although it has some of its own flows, still it is more successful than other exams to separate best from the rest. As Manish Verma, who runs coaching class (http://www.123iitjee.com/iitjee_blog.htm)states: "apparently, the exam tests maths, physics and chemistry, but in reality, it measures speed, strategy and stamina of the candidate".

    2) Kamal sinha, a graduate of IIT-B and critique of IIT system and IIT-JEE (http://www.kamalsinha.com/iit/iit-jee.html) states that JEE has not achieved much.

    3)JEE is a problem solving exam, and it has been reported that the JEE rank roughly equates IQ of a student, with top few hundreds can be classified as highly intelligent. Foreign companies ask for JEE rank of a candidate, and Google selects only from top 50 AIR for its specialized work, offering pay of USA for work in India.

    4) The first entrance exam was a modest exam for IIT-KGP in 1952. Earlier each IIT had its own separate exam, and to reduce hardship of students, a common exam, called IIT-JEE came into existence in late sixties. In our time (1972), the problems were quite complex and there was also a paper of English, due to which almost all students were from metros offering coaching classes. Even today, a newspaper report states quoting IIT-K faculty that 90% of the children of IIT-K faculty could not clear the JEE becuase of lack of coaching class in the city.

    5) The JEE format was changed to a more complex form in late fifties as more IITs were coming up and competition for bright students became intense. It also become tough as more exotic and mathematically oriented courses (electronic, aeronautics and naval architecture) were introduced, requiring students with sound fundamental knowledge and grasping power.

    6) The JEE is difficult than AIEEE and it is not possible to correlate the two exams, unless some common dummy questions (not counted for marks) are inserted in both the exams. My random check on ranks obtained by coaching class students for both the exams found that a typical student will receive a lower JEE rank than AIEEE rank, although more students appear in AIEEE. I found that a rank of 1000 in JEE corresponds roughly to 300 AIEEE.

    7) It is not possible to combine the JEE and AIEEE, due to centuries old caste system percolating in our IIT system. The JEE board vigorously defends the exam (like a snake guarding a pearl in a ruined house). It has denied entry of Indian Institute of Foundry and Forging to JEE, removed Bombay Marine Engineering School and threatened Indian School of Mines in 2004 to delink. Roorkie got bad treatment from rest of the IITs because its four batches got IIT tag when it became IIT in 2002. Also, IIT would not like to depart from JEE.

    Sorry for the long letter, which the topic demanded.

    Thanks,

    Yogesh Upadhyaya
    New Jersey, USA

  4. barbarindian said...

    You missed out bias against Dalits and OBCs.