Saturday, September 26, 2009

My take on CAT '09

Yes I know that it has been pretty late since I last posted on this blog, and that its just a few weeks away from the D-day (or the D-week or what-ever) but still I thought its never late for a good thing, so lets try to have a look at what can we expect from the online format of the CAT.

A few things which I want to clearly state before going into the main body of the post, all the things which I am going to write are just my views and predictions, and they may or may not be true, I am not going to write things like the number of questions one can expect in CAT but the type of questions which have more probability of appearing in CAT.

QA :

CBT or PBT (paper based test) , this section should be consisting of similar questions, testing the same set of skills and so it shall need similar approach as theconventional CAT exams (as if there is something like conventional CAT :O).

The point to emphasize here is, even though the total number of questions may decrease (most probably it will) and almost all the coaching institutes are emphasizing that one has to be good in all topics in order to do good in the exam, it isn’t so.

According to observation of previous year CATs one thing is for sure, 40 % marks in any section is sure to get you 98+ percentile, this effectively means you need to do 8 out of 20 (or so) questions right. So why do coaching institutes emphasize on studying all topics, because you never know questions from which section would be easier, the paper may have easier questions from the section you have left while difficult ones from your areas of comfort, now here is a trade-off : you either work on all the topics, so that you can do the sitters from all the areas well or you may decide to be so good in your favorite topics that you can solve the questions, come what may.


DI/LR:

DI is the one section which will probably get affected most.

Reasons:
i.) Fewer number of questions means either there will be fewer sets or fewer questions per set, while the former will make question selection tougher and more important, the later will make sets less worthy and hence you will need to do more sets in the same amount of time for having our desired percentile.
ii.) Aspirants used to fill the tables and other figures with data while attempting DI/LR questions, this facility won’t be available now + It isn’t easy to look at the screen each time you need a data.

No matter how hard this problem seems but it can be solved by one remedy “PRACITCE”.

VA/RC :

More probably then not this time CAT will be emphasizing less on RC questions then on other questions, questions on usage/jumbled up sentences/sentence correction are likely to form major section of the paper, short passages with lesser number of questions than usual may be a surprise package for many.

Suggestions to improve in this section: emphasize less on reading speed and more on comprehension/analysis of the read text, don’t depend too much on techniques like reading the question first and then trying to find out the key-word in the passage but try to actually understand the passage and then answer the questions, if possible work on vocabulary, it might be of help, get acquainted with grammar rules etc.

Some more points to note.

==> Data sufficiency questions may be back in fashion, both in QA and in DI.
==> Less emphasis on lengthy table based DIs, more on crisp logical reasoning questions and a few pie-chart or other figure based DIs, which have many small and related figures per set.
==> Analytical reasoning questions will be in something to look for.
==>Be comfortable and confident, why ??? let me take an example of QA.

There would be somewhere around 20 questions of QA that you will have to do in 45 minutes.
Now lets say you need 4 minutes to look at your watch, adjust the mouse, scratch your neck and other such inevitable things.
Now lets take the following assumptions.
i.) you read all the questions.
ii.) you solve 10 questions, out of which you solve 8 correctly rest 2 you either solve in-correctly or you leave them in-between coz they seem to be lengthy.
iii.) You have an average reading speed, and you take close to 80 seconds to read a question and realize whether you should try it.

Now usually there would be at-least 2-3 questions, which would be from the topics you dread, so these questions you will leave even without wasting reading them and wasting much time, lets say taking 30 minutes per question. So you would have seen 3 questions in 90seconds (Wow, that’s quick !!!) + you will take close to 560 seconds in reading the questions you won’t attempt.

So in total you have more than 30 minutes to attempt the10 questions, out of which 4-5 will be sitters which can be easily solved in about 90 seconds each. Attempting 4such questions would mean you have 24 minutes for rest 6 questions that means whopping 4 minutes per question, and these are the easier questions from the topics you are comfortable in.

Now doesn’t it seem obvious that more than 98 percent of people should get a sectional of 98 percentile plus in QA :P

Sunday, September 6, 2009

Role of Question Selection in CAT ...

Note : This post is the same as the reply given by by me at pg, except the last para which concentrate on the role of question selection in Computer based test.


Question Selection always plays a big part in competitions like CAT, and though it largely depends upon a no. of factors like the paper format, relative difficulty of questions and all, but the most important factor is YOU.

You have to realize that what are the questions that you are comfortable with, what are the questions that you can solve in less time, for some guys like RCs are the only thing they are confident about in VA, for some-other guys RCs may be dreading while Vocab a nemesis.

But then you have to understand, that at times you have to make a trade-off and leave something you are good at just in order to make sure actually get the thing that matters most, a decent(and balanced) percentile.

Now though it depends on person to person, but still I would give u some points which might help you, please understand that these are my personal opinions and hence try them in mocks to see whether they are actually suitable for you.

i.) Don't make any pre-assumption about the paper like I will solve 15 questions in quant, rather chill out n try to find questions which look familiar to you, in the sense that either you have solved/seen such question earlier or the concept involved looks easier.

ii.) If there is any question involving terms such as 'n' or 'x' give yourself 2 minutes to try to put various values of the variable and option check.

iii.) Don't give a LR puzzle more than 5 minutes unless you actually get an idea on where it is leading to, or you have no better question to solve (LR puzzles are kind of addictive particularly for engineers and guys who love puzzles, but then its a exam).

iv.) Make use of bookmark : if u realize that a question looks familiar but you aren't able to crack in a few minutes bookmark it and solve other questions, come back to it once you feel like you have solved all the easier questions.

v.) If you aren't a voracious reader chose passage in which the questions are of factual nature eg: which of the following is the reason cited by the author, or which of the following is the reason provided .....
These kind of questions can be solved by just skimming through the passage, and then looking for the keyword(of the question) in the passage.
Avoid philosophical RCs and questions like, what does author think OR what does author implies.

And now, lets see how question selection differs from a paper based CAT to a cbt CAT.

You must have seen people telling that you should do the questions in two go, one of x min and the other y min, In a cbt exam this could be pretty much time-taking, its not easy to bookmark 10 questions and then you keep searching the questions which you have bookmarked, plus if you guys have appeared in online mock-cats, it takes some time for the questions to load, so in my opinion the strategy needs to be changed a little bit.

Go for all the questions that you think, can be solved in under 2 minutes, bookmark questions only if you know they are easy but you are missing some key point OR if the question is easy but involves lot of calculations avoid using excessive bookmarking. Go to the other sections, do them and then try to analyze your sectional performance and accordingly decide to go again into which section, go and solve the bookmarked questions in the section.

Best of luck !!!