The PaGaLGuY.com Business School Rankings

[grondmaster]

Originally Posted by pagalguy

Grond - I agree on the first issue. This should be sorted out in an hour or so. Escaped beta testing.

That's OK. That's why we have crowdsourcing, right

Originally Posted by pagalguy

The 2nd issue I'll wait till Rajat comes in. Were you able to complete the survey?

Unphorchunately not.... Haven't tried a second time... Maybe tonight...

Meanwhile... some discussion from my side about Apurv's take...

Originally Posted by Apurv

Faculty that does make a difference in b-schools, and any b-school student would testify to that, is that which makes for a great classroom experience. Measuring that with hard data is not possible. Perhaps a perception survey of best profs in each b-school answered by the students, the results of which are integrated into a larger perception survey.

My two pfennigs:
Almost everyone in my school, in my batch & the junior, will testify to the very practicality of the teachings of our stats prof. He's not a full time prof., rather a consulting expert, who comes and teaches us. His theory lectures are few and far between. All others are proper case based, where we keep doing stats sums. He also detests the normal training and lecture fundae that the profs keep pfaffing about, rather stressing on practical knowledge than theoretical.

In our summer internships, a specific method, which we now take for granted as a basic stats tool, was touted as an advanced technique by other bschool students to me & my batchmates. Not that they didn't learn things we did... rather, our prof made us lose the fear we had of stats by showing us what exactly we were doing. That's why we weren't so enamoured by them.

Each Bschool will have some such profs, and a lot of the other variety. Faculty has to be equated to the knowledge they pass on to the students. Does it matter if that poor guy is just a normal postgraduate, or holds multiple PhDs.????

Originally Posted by Apurv

2. Placements - Cut-throat competition between b-schools has led to a situation where placements data shared by b-schools is suspect. We have seen deep enough into placements processes in b-schools to know that distorting facts to project higher average salaries is a game that has become a norm. Inflating the CTC, averaging by combining international salaries with domestic ones, over-highlighting certain data to hide some other data... the methods are as many as there are schools. The companies don't mid either because of the PR they receive in the process. We think that in such a scenario, comparing b-schools through hard placement data will end up being an exercise in vain. In fact it would amount to construct a second lie over another.

Agreed... In fact, seeing the wars that happen everytime placement data is put up on PG, I was debating whether to ask Allwin for a blanket ban on placement discussions, just as we have on BSchool comparisions and the full-page-ad-wala institute.

Originally Posted by Apurv

4. Intellectual Capital - Not conclusively quantifiable except in maybe 10 b-schools.

Can someone care to define this parameter in a quantifiable manner? Such a vague thingy is dangerous for scoring, and is truly debatable as well. And i don't think we can even get a consensus on the 10 schools you're talking about...

Originally Posted by Apurv

Moreover, hard data, unless primary and verified, is questionable. Objective rankings that exist today rely on questionnaires sent to b-schools, the answers of which are taken on face value. Visiting even 100 schools for verification is so humongous a task that nobody has yet taken it up.

Erm... and then some bschools fight the rankings they hate by countering a higher moral ground - they say that they did not participate or fill up the questionnaire, and the ranking entity has fudged the data based on out-of-date material...

Regards,
Grondmaster

[pagalguy]

Three of the issues have been fixed. Thanks to Rajat's relentless tweaking of the system. Some of you were able to take the ranking twice, in which case due to a timeout issue your first submission was not recorded. We have now fixed it. Go ahead and submit the survey!

Secondly, the MBA Alumni list has been expanded. That was an oversight on our part. Now all the schools have been included in that list

The Name of the states have been alphabetized so that you can breeze through the form much faster!

Thanks for the patience folks

[capreal26]

Just filled up the survey.. first up.. this is one helluva positive application of wisdom of the PG crowds. Team behind this concept & execution deserves an applause even before the results are out. . Now lemme come to the usual part, viz easy - critical analysis
I read through the methodology of rankings & hopefully understood it. Its a simple elimination round ranking, so to say where the one who emerges on the top has to be checked in all of its questions & so on.. Also, read Apruv's post on the reasons behind not going with the data based ranking methodology. Agree with that too.

One thing this will conclusively do is - layout the rankings based on perceptions. So, if another one of India's best selling business magazine comes out with a ranking which shocks 95% of readers, we know where to go.

But the fact that those perceptions are also based on a long history of half truths which have been fed to the responders by the so called mainstream media, network of friends & other sources of info. & We can reasonably assume that there would be no one of the list of responders who would have even basic information about all of the colleges listed there. So, when you are asking me to pick a XYZ based in Nagpur with ABC based in Coimbatore, my responses are near random and can change the next time I take the survey. So, how do you plan to factor for this error. Thats a separate discussion.

Second thing - which is more of an idea to expand the same survey -once you have the initial results (which is your ranking of 1-70), is it possible to break this up into 2 (or maybe 3) buckets of say 1-20 & 21-70 and run the process again. [I know the team is not gonna like this armchair statistician's ideas but hey who carez ].
Twofold reasoning for this -

1. Most of the money is in the outliers, i.e. its the top 20 list which will be the object of interest. I dont think it matters much whether I come 44th or 56th.
2. While its always an easy answer when you compare a top institute to an unknown insti, the real perceptions will be uncovered when you are asked to compare IIM X v/s IIM Y.

btw, this means the above is a separate survey which should not be taken by the people who have already taken the first one.

ideas/abuses ?

Am sure all these things would be under consideration from the team behind the project, but since I am at it, lemme go the whole hog.

What about breaking the perceptions down by areas mentioned by Apurv. So, questions would be like - which of these schools do you think rates higher on Infrastructure? or which school's faculty do you think rocks more? So, your end results are in the form of rankings on multiple areas. All still based on perceptions but much more detailed. & should be more fun I guess.

& finally, after the team succeeds on perception based results, it should be time to get into preferences. Which is when the whole structure & objectives of the game would change. Broadly speaking, the objectives of uncovering the B Schooling preferences of Indian crowd would be to rank various parameters (some of the ones listed by Apurv) in the order of importance.

If you ask anyone where do they want to do their MBA, am sure anybody would say a B School which has the average salary of 150K USD, has the fee structure of FMS, is based in downtown Manhattan, gets its faculty from H/S/W/top 10, has the best pub on campus, has the infrastructure of ISB (), is an hours drive from their home, has a 50-50 male-female ratio & yes, btw, doesnt focus much on grades. But I dont think we have a school like this as yet. Long story short - you cant have the world. So, what are the preferences (which are completely different from perceptions - currently being measured) of Indian crowd. Do you want maximum number of IB jobs or a support system to start your company on campus, etc etc. That I am sure can be the point of arrival of this initiative.

Then if we can tackle the problem of unreliable data, guess what, we can actually match the preferences to the best school available in Indian market for various sets of people. Dont I sound like an MBA myself.

so, ppl what do ya think. am hooked on to this for more action hopefully.

[pagalguy]

Originally Posted by capreal26

Just filled up the survey.. first up.. this is one helluva positive application of wisdom of the PG crowds. Team behind this concept & execution deserves an applause even before the results are out. .

Thankoo thankooo .. a major part of the concept and analysis was done by our summer interns! Talk about doing substantial work at PG

But the fact that those perceptions are also based on a long history of half truths which have been fed to the responders by the so called mainstream media, network of friends & other sources of info. & We can reasonably assume that there would be no one of the list of responders who would have even basic information about all of the colleges listed there. So, when you are asking me to pick a XYZ based in Nagpur with ABC based in Coimbatore, my responses are near random and can change the next time I take the survey. So, how do you plan to factor for this error. Thats a separate discussion.

Very true. But at the same time we give people a lot less credit than they should be given. A lot of people discount the salaries, a lot of the hype gets discounted and as we say it is the knowledge of the crowds that finally help us build a ranking that is valid nationwide. Your point about Nagpur v/s Coimbatore is very valid and is something we want to capture. People in the North may be more familiar with schools in North and the people in the South may be more familiar with the ones in South. Overall - our insights and local knowledge is averaged out . However, even though you consider the deal to be random, I don't think it is completely so. Even if it is random, may be that is what a student's call may be on that school. Random. You may be indifferent between a school ranked 55/56/57. So our knowledge of the crowd concept works out just fine.

2. While its always an easy answer when you compare a top institute to an unknown insti, the real perceptions will be uncovered when you are asked to compare IIM X v/s IIM Y.

But we are absolutely doing that. In our current process, every IIM is being compared to every IIM and every school out there. So why should one redo the process again? Wouldn't it be redundant?

What about breaking the perceptions down by areas mentioned by Apurv. So, questions would be like - which of these schools do you think rates higher on Infrastructure? or which school's faculty do you think rocks more? So, your end results are in the form of rankings on multiple areas. All still based on perceptions but much more detailed. & should be more fun I guess.

Its like asking about who would you marry. Factors could be numerous, and individually they don't really matter as much. The final call you make is a triangulation of all the factors and that is what we want to capture. But I see the point you are making and yes we are working on ways to further slice the data. At some point we would be able to tell you how a regional ranking looks. You did select your state when you started the process didn't you?

At some point, we might even be able to tell you precisely what people in the south think about schools in the north and vice versa and beyond. May be there is a difference between how men and women think about business schools. We could bring that up in the rankings at some time with enough data

Then if we can tackle the problem of unreliable data, guess what, we can actually match the preferences to the best school available in Indian market for various sets of people. Dont I sound like an MBA myself.

That is an application of the data we might be able to generate at some time. Heck, it sounds cool

This would be just the begining and the insights might stun people. We are seeing major trends and also we are now getting to see first hand the elasticity between different schools which is something no one could have measured before

[deepakjnair]

The link is not working....After i fill in the form ,it shows a problem page..

I have attached the screenshot.. Any issues???

[QuintEssence]

Originally Posted by Apurv

<<snipped>>

Appending this post to PG's first post would make a lot of sense for the first-time reader

Ranking methodology, very neat. A is taller than B and C is smaller than B but a large number of respondents would be required for the data-set to become crunchable and relevant, no ? how about sending a PG.com notifier email to all users ?

[pagalguy]

Originally Posted by QuintEssence

Appending this post to PG's first post would make a lot of sense for the first-time reader

Ranking methodology, very neat. A is taller than B and C is smaller than B but a large number of respondents would be required for the data-set to become crunchable and relevant, no ? how about sending a PG.com notifier email to all users ?

1) Post appended!
2) Very true, but the response has been phenomenal and we are now getting started. We have collected more data in 12 hours of the launch than MR agencies could have collected in 6 months . Emails, pluggage and all to start today so that we can get it to be as accurate as we can. Meanwhile, share it with your friends, old classmates and alums. Would love to see the rankings reach a wide spectrum.

[floyd&vodka]

Originally Posted by pagalguy

1) We have collected more data in 12 hours of the launch than MR agencies could have collected in 6 months .

When you guys made this survey, you must have made an estimate of the minimum number of respondents (the sample size) you would need to be able to draw logical and fair conclusions out of the responses.

Can you plz let us know, what number are you looking at for the minimum sample size of this survey?

P.S: I am assuming that at the end of the survey, you would be revealing the actual number of respondents in the various segments who took part in this survey. However, I am asking the above question to understand the methodology better.

[arunk8186]

first of all congrats to the PG team for havin taken ths initiative.it willll definitely help us MBA aspirants a lotttttttttttttttt!!!!

BTW can someone plz tel wen the poll results wil b put up....after gettin 1000 responses or wat??..

Thank uuuuuuuu

[pagalguy]

Originally Posted by floyd&vodka

When you guys made this survey, you must have made an estimate of the minimum number of respondents (the sample size) you would need to be able to draw logical and fair conclusions out of the responses.

Yes, and it is both a quantitative and a qualitative answer. Quantitatively, we need 57 people to answer, so that they would have answered 57 x 50 = 2850 questions which would enable us to build a ranking because every school will have been compared against every other school.

Qualitatively, we want to get as many multiples of that set to ensure that we have appropriate representation nationwide.

Can you plz let us know, what number are you looking at for the minimum sample size of this survey?

No, we haven't decided on a minimum number. I mean if we had less than 57 people respond - we wouldn't even be a rank. But we have already gotten tons more and are continuing to get more.