Tuesday, June 23, 2009

Publications in IIT/IISc

The number of publications from 2004 to 2008 has grown in all IITs and IISc. Here is the trend.




One can clearly see that IISc has the largest number of publications but the number of publications has increased significantly in other IITs. For example, the ratio of the number of publications in 2008 to the number of publications in 2004 is 1.2 for IISc, 1.7 for IIT-M and 2.0 for IIT-Roorkee.

I will post separately on the quality of publications but whatever parameters one takes: number of citations, number of citations per paper, h-index, number of papers in journals with impact factor of more than 5 etc., in each category, IISc leads the others. However, when one compares this output to that of the top 30 universities in USA, the output is still low. Therefore, this shows that there is a scope for lot of improvement for all institutions in India.

12 comments:

Venkatesh said...

So, what????

It is well known that we can not compare the IISc with IITs. Look at teaching loads that IITs have!!.. I am much appreciative of the guys at IIT who are still competitive in terms of research output while coping with the teaching. Shame to IISc where they are not able show greater output while they have lot of time and money at their helm and moreover the teaching load is minuscule when compared to that of IITs.

Kudos to IITs and the faculty there.

Anonymous said...

Professor Giri is only trying to say that there is scope for improvement. Read the last sentence of the post.

IITs have no more teaching load than any of the top 30 universities in USA.

Anonymous said...

hmmm.. Mr. Anonymous, the two last sentences were not there when I read the post!. That lead me to write my first response. Probably, Prof. Giri hadn't completed it. I was one of those early birds who visited before it was complete!

In any case, the support the USA universities get and the atmosphere they create for a faculty isn't the same either. Socio-economic status of an individual also does matter for the output man!.

My opinion is that, when we make comparison.. we should only be comparing the institutes of same stature! nothing more nothing less.

Niket said...

Venkatesh,
The graph shows the total number of publications. A more appropriate figure would be number of publications per faculty. I guess IITs have twice the number of faculty than IISc.

The comment "while they have lot of time and money" is incorrect. I have been asked by my aunt why I don't sleep before midnight when I have just one three credit course to teach per semester.

The number of publications per faculty in either IITs or IISc is not comparable to that in first rung American universities. The comparison is apt because it is we ourselves who use favourably compare IITs to these universities (I recall how the 60-minutes piece on IIT was widely circulated).

Secondly, shouldn't we compare best in India with best around the world?

This comparison does not diminish the fact that we have come a long way in the last 50 years; it underlines the fact that there is a room for improvement; that improve we must.

Anonymous (5:34 pm),
The statement "IITs have no more teaching load..." hides various other things required for productive research. First, our department (ChE, IIT-M) averages almost 5 credit hours per faculty per semester; its 3 in top-tier US universities. Second, we spend a whole lot of time doing mundane things due to lack of technical support. Just to give an example, there is a faculty in charge of the website, faculty in charge of maintainence unit, faculty in charge of estate... that takes time and effort. Finally, the quality of students entering MS/PhD programmes in the two countries is not comparable. US gets best in the world; we seldom get best in India.

All this is not meant as an excuse. Its a reminder that we have to meter our expectations based on the ground realities. Still, as people like Giri have shown, it is very much possible to run a productive research program in India.

Anonymous said...

Niket,

I don't think all IITs have twice the no. of faculty as IISc - this might be true of IIT KGP, while Kanpur, Bombay and Delhi are smaller. And the teaching loads, at least in terms of the class sizes, at IITs are much larger - no debate on that. Here again, comparison with US is bogus - sinch much of the work is done by TAs (assignments, marking etc), while I don't think IITs have evolved such a serious TA system.

I fully agree with the third comment on the support and the atmosphere the faculty get in US.

Although we compare IITs/IISc with top world univs, this is not apt, given the ground realities. We should hold on to some more time, till a fair playground emerges (if at all it happens)

kumar

Anonymous said...

Another point - I think IISc's lead in publications is due to the strength of its science departments which have been traditionally very strong. Plus, Science disciplines usually publish way more than engg depts. It would be interesting to see the break up, science and engineering -wise of IISc and comparison of engineering publication with those of IITs.

kumar

Giri@iisc said...

The number of faculty in IITs and IISc is comparable. For example, the teaching faculty strength of IISc is 87 asst prof, 87 assoc prof and 167 prof. These numbers are not significantly lower than that of IITs, which are in the range of 400 to 500.

It is also incorrect to state that IISc has bigger science departments. The size of chemistry and physics departments in IISc and other IITs are comparable. Of course, IISc has a large biological sciences program of 40 faculty which IITs lack.

Previously, I had published an analysis on the chemical engineering publications in IITs and IISc that included the number of faculty i.e., papers per faculty, citations per faculty

http://www.ias.ac.in/currsci/may252008/1265.pdf

A similar analysis has been conducted for all departments (chemistry, physics, elec engg etc.) and that paper is under review. So, to clarify, I have indeed done that kind of analysis.

My aim is not compare it against the top 10 of the world directly but more to examine where the best departments and institutes in India stand with respect to the top universities in the world including developing countries. That's why my article above also compares it against the top universities in china, korea etc.

Regarding the comment,
"Shame to IISc where they are not able show greater output while they have lot of time and money at their helm" it is unfortunate you think so. The amount of money for research in IISc and IITs are about the same. If you want me to discuss the plan and non-plan grant of each IIT and IISc, we can do that or one can look at the budget papers of the government, available on the web. Go to the mhrd website or the finance website of the government and get the statistics. Further, IITs have a large heritage alumni grant which IISc does not.

The number of administrative and technical staff in IISc is roughly 1/2 of that of IIT. Some of the faculty like me are loaded with some much administrative work that it takes 2-4 hours a day doing this work. I am not complaining but only telling you we do not have "a lot of time"

Yes, the teaching load of IITs is twice than that of IISc (six credits for a larger class per semester versus three credits for a smaller class per semester) and it is commendable that IIT faculty are productive but there is no shame in being in IISc.

To repeat, I am not trying to compare IITs with IISc, but only top institutions with comparable sizes with one another and with the rest in the world. If one removes the criteria of comparable sizes, institutions like TIFR and NCBS are doing better than IISc in terms of research output per faculty.

Thanks

Giridhar

Anonymous said...

I think that the very concept of making publications a metric to quantify the success of a university is wrong. There are a lot of bogus papers that do not add any value to society and the knowledge base.

Sure, if I get a piece of a new polymer-based substrate, shine light on it and measure its photo current, I might be sitting on the mother-lode of the next generation of solar cells. But, more often than not, people would change the composition of the polymer by 10%, shine UV/IR light on it and measure under different temperatures. There you go, three different papers on "compositional variance", "wavelength sensitivity" and "Thermal robustness". :-) Then, have a few of your buddies refer to your paper and you refer to theirs and the citation index goes up.

The real product of academia is a student and the contributions she makes in future. Is there a metric that quantifies it? If IISc delivers a very inexpensive disease monitoring kit that revolutionizes rural healthcare in India, what metric measures it? Specially, if the only innovation is in cost and so, it cannot be published in a "famous" journal.

I think we should make "technology transfer" a key metric in assessing the performance of IITs and IISc. We may not compare well to Harvard and Stanford, but was that the reason Sir J N Tata established the school?

Bharat

Giri@iisc said...

Dear Bharat,

Nobody said that publications or citations are the only criteria in assessing the institutions. I have indeed done the analysis including the number of patents filed, technology transferred, low cost devices developed etc. But these numbers are too small to be compared or talked about.

Giridhar

Anonymous said...

An interesting observation: it seems that IIT-Kharagpur has maintained a considerable gap between itself and other IITs for the last few years.

I propose to the MHRD mandarins in Delhi that all the new IITs are established in the middle of nowhere with pathetic medical facilities and a stone's throw from areas of Maoist insurgency. This seems to have a positive impact on the quality of research ;-)

Watch out Prof. Giri, as soon as they establish IIT-Kashmir, IISc will no longer be the research leader in India.

revathi said...

IIScs perhaps have more full time researchers than IIT which means that they can be more productive. On the other hand, IITs have more student work force who carry out short term projects that can be highly productive. So the result is not surprising.

Giri@iisc said...

IISc does not have full time researchers. Faculty who do research also teach.

I doubt quality publications can result from short term projects. To claim IITs produce publications through short term projects is not correct. Many IITs graduate more Ph.D's than IISc.