CAT 2006: Formulae for Error-correction

[mufasa]

Let me jump into the bandwagon and increase my count...Not that IIMs would follow my suggestions..
1) Have different cut offs for different sets
2) quant :
luggage wala question: Those who have got the right one correct, give them 8 marks
Venn diagram wala question: scrap it
3) Eng : Those who have attempted 3/5 in that RC, give them marks for the 3 answer option wala question..
4) LR : havent seen the question, so cant comment
5) Call those who are missing the cut offs by 1-2 questions....

Did I miss any of the wrong questions ?
Unlike the retest wala thread, I have considered the people on the other side as well while proposing a solution.....

[mufasa]

Originally Posted by gaurav2cool

look u were quick 2 close d thread n now startin a new thread 4 error correction
i dnt think opening such a thread would b of any help because IIM'S will do
what is feasible 4 them i dnt think they need any suggestions i m sure they will come
out with it in jan so no need 4 this

Don't you know that IIMs have outsourced this issue to us....The puys have been given the responsibilty to find a solution..where have you been so long?

[logic]

Originally Posted by tatimatla

My Formula:
Quant questions that were wrong: Scrap. If a person had attempted the first question (in the set of 2 qs)...award him +8.
DI question that had 2 answers: Deem both correct.
VA: No Comments.

thats the right way. i agree with u on that.

[gpriyanshu]

Junta, please remember lowering the cutoffs might not be as fair an option because cat scores are used by institutes other than the iims as well. While it might be a good idea that iims call a few more candidates for gd/pi and reduce weightage given to cat score in final selections, we cannot expect all the institutes using cat score to lower their respective cutoffs.

[mufasa]

Originally Posted by gpriyanshu

Junta, please remember lowering the cutoffs might not be as fair an option because cat scores are used by institutes other than the iims as well. While it might be a good idea that iims call a few more candidates for gd/pi and reduce weightage given to cat score in final selections, we cannot expect all the institutes using cat score to lower their respective cutoffs.

It is not fair if the paper itself is error free...If not then this becomes a good option to include those who have been affected by the errors... We are trying to find a viable option to a retest solution...I agree that there might be a gulf between the 99.5 and above guys wrt others...but between [98-99.5), I don't think that there is alot of difference...it's just a question or two..So the point of diluting the quality does not hold..
Regarding the other colleges, well I don't think they would be complaining, coz I feel that IIMs will come out with a good solution....

[on_da_rox]

Originally Posted by tatimatla

My Formula:
Quant questions that were wrong: Scrap. If a person had attempted the first question (in the set of 2 qs)...award him +8.
DI question that had 2 answers: Deem both correct.
VA: No Comments.

nah,i dont think so...
of course,the 1st question may not be scrapped,but awarding +8 is too unrealistic... tho reasonably,the whole caselet should be scrapped,plenty of people (including yours truly,a lots of friends) left both the questions because of the contradiction of the 2nd questions....
and both the answers should be treated as correct for the DI q

[on_da_rox]

Originally Posted by buzzeG

scrap the wrong Q's.. in the 2 set Q scrap the second one if the answa is anythin othr than 30 and full marks if answa is 30 (people mite hav arrived at the answer of 15 and not finding it there, considering that there were 2 ppl travelling, marked 30)

DI both answers are rite..

VA of course there is no error only ambiguity (except for the partially printed Q in one set which naturally has to be scrapped)

phew!! quite some altering huh... like cuttin a nice evenin gown to miniskirt.. i feel thats the way out now.. rather than pulling it all off and puttin on a new one

hehehe....funny choice of words. ...pretty suggestive too....
just hope that the CAT retains its modesty & just restricts itself to "cutting the nightgown to a miniskirt" to put it in your words

adios

[gpriyanshu]

Originally Posted by mufasa

It is not fair if the paper itself is error free...If not then this becomes a good option to include those who have been affected by the errors... We are trying to find a viable option to a retest solution...I agree that there might be a gulf between the 99.5 and above guys wrt others...but between [98-99.5), I don't think that there is alot of difference...it's just a question or two..So the point of diluting the quality does not hold..
Regarding the other colleges, well I don't think they would be complaining, coz I feel that IIMs will come out with a good solution....

well, u seem to have missed the gist of my post. I have always felt and said that beyond a few say 50 candidates, all atleast another arnd 10,000 odd ppl are almost of the same level, its just luck on that particular day. I never and can never make the point of dilution of quality. so giivng gd/pi calls to more ppl and also lowering relative importance to cat in final selection might help in undoing some of the injustice.

However, my concern simply related to other institutes who cannot all be expected to lower cutoffs. So, a simplistic solution like giving more gd/pi calls might not work and a formula needs to be arrived at in the score n percentile calculation stage itself, which i'm sure iims wud suitably do. ofcourse, it can or rather it should be supplemented by more calls.

[mufasa]

Originally Posted by on_da_rox

nah,i dont think so...
of course,the 1st question may not be scrapped,but awarding +8 is too unrealistic... tho reasonably,the whole caselet should be scrapped,plenty of people (including yours truly,a lots of friends) left both the questions because of the contradiction of the 2nd questions....
and both the answers should be treated as correct for the DI q

Nah, here is my reason for awarding the 8 marks....
The questions invloved solving 2-3 simultaneous equations in order to arrive at the answers...If one solves them then he gets both answers...So provided someone has marked one answer correct, it can be assumed without doubt that he has solved the equations...The questions were not unrelated, that's why this logic holds....
Regarding people like you have mentioned... What this approach does is that it does not harm those who have sloved the set...and at the same time if calls are given to those who are missing the cut off by 1-2 questions, it also ensures that they are not missing out because of the error...
So you see both sides are covered...... Comments?

[mufasa]

Originally Posted by gpriyanshu

well, u seem to have missed the gist of my post. I have always felt and said that beyond a few say 50 candidates, all atleast another arnd 10,000 odd ppl are almost of the same level, its just luck on that particular day. I never and can never make the point of dilution of quality. so giivng gd/pi calls to more ppl and also lowering relative importance to cat in final selection might help in undoing some of the injustice.

You did not mention it, but I sort of inferred it... Seems like I was off the mark..

However, my concern simply related to other institutes who cannot all be expected to lower cutoffs. So, a simplistic solution like giving more gd/pi calls might not work and a formula needs to be arrived at in the score n percentile calculation stage itself, which i'm sure iims wud suitably do. ofcourse, it can or rather it should be supplemented by more calls.

About other colleges- to tell you truly I havent yet considered that point until you brought it up ..that's why I did not talk much about it in the previous post also.... But even then, I am sure I can come up with a solution... that would be better than a retest..