Recognizing Shortcuts

[mgfeener]

Hi All-

I have been studying for Exam P for about 6 weeks now by going through The Infinite Actuary course.

I am about 2/3 through the material and have worked a fair amount of practice problems.

I am consistently answering all the questions correctly, but it often takes me about 15-20 minutes per problem whereas Dave can solve them in 3-4. Also, I am frequently using strategies different than Dave does (many of my strategies involve inserting Calculus where it's really not necessary....)

Does anyone have any pointers for learning to solve these problems a little bit quicker? Is this all part of the normal growing pains of learning the material? I have no doubt I can pass this exam given 9 hours to do so, but I'm not sure exactly how to go about studying to complete the test in the allotted 3 hours!

Any feedback is appreciated.

Regards,
mgf

[Actuarialsuck]

Do problems, then do more problems, then just for fun... do more problems. If you have problems left, do all of those too.

[aaronhsmith]

I think having good test strategy will help you when you sit. Go through and only answer questions you can do easily and quickly (let's say in under 3 minutes). As you go through, mark the one's that you can do easily, but will take more time. Those are the one's you'll do on your second pass through. Cap yourself at 6 minutes each (or so) for those. With whatever time you have left, you can go back through and work on some of the harder problems. On your third pass through, you shouldn't have any more than 10 questions left if you're well-prepared enough.

Guo's manual talks about setting up each problem with a "Three Minute Script" that you can follow. Also, if you're looking for shortcuts, his book is good for that. Lots of tips to save precious time on the exam.

All the best.

[JShreib]

Hi All-

I have been studying for Exam P for about 6 weeks now by going through The Infinite Actuary course.

I am about 2/3 through the material and have worked a fair amount of practice problems.

I am consistently answering all the questions correctly, but it often takes me about 15-20 minutes per problem whereas Dave can solve them in 3-4. Also, I am frequently using strategies different than Dave does (many of my strategies involve inserting Calculus where it's really not necessary....)

Does anyone have any pointers for learning to solve these problems a little bit quicker? Is this all part of the normal growing pains of learning the material? I have no doubt I can pass this exam given 9 hours to do so, but I'm not sure exactly how to go about studying to complete the test in the allotted 3 hours!

Any feedback is appreciated.

Regards,
mgf


I'm assuming you are not sitting for the exam in September? If not, I wouldn't worry about this yet. When first learning the material a lot of the problems will take longer than the "6 minute benchmark". This is good since you're learning as you struggle through problems whether you realize it or not.

I felt this way on FM more so than P but once you start taking practice exams you'll most likely realize that the time limit isn't as big of a deal as you think it is now.

If you notice that you are unnecessarily using calculus just try to learn from the solutions provided by TIA. But until you start taking practice tests, strictly focus on the material and not on your speed.

Also keep in mind that there will be "easy" problems on the exam that will take much less than 6 minutes. This will leave more time for the more difficult problems and you will be way more prepared at this point.

[mathmajor]

When I first start practicing from a mixed bag of problems, I'm lucky to even get a feasible final answer. After a while things get easier... they can only ask problems in so many ways. "Been there, done that!"

[bifodus]

There seem to be three steps to solving these problems.

The first step involves understanding the concept going back to the definitions. For example, if a problem is asking for the variance then you want to, in theory, be able to derive the answer by going all the way back to the definition of variance and then crunching your way to a solution. This would take time, but it's where you want to start.

The second step is to be able to use knowledge from previous calculations to shorten the process dictated by step one. If I need the variance of an exponential random variable then I can shorten the process by using my acquired knowledge that all I need to do is square the mean. Finding the mean becomes sort of a sub-problem (you can recursively go back to the first step).

By the end of step two you have a plan, and so step three would be to carry out that plan. This means crunching, and it should be purely mechanical.

It's clear that the step that's slowing you down is step two, and the only way to speed it up is to build up your bank of completed problems. This will help you to categorize problems by certain characteristics. For example, if I see a time span problem then suddenly a bunch of information pops in my head regarding the relationships between Poisson, exponential, and gamma distributions. This type of thing allows for a quick derivation of a quick plan.

[mgfeener]

This is all very great insight guys. Thanks a lot.

I think I may have been a little naive when I said "I've solved a fair amount of problems". The more I read around the forum, the more it is becoming clear that I still have many, many work problems to work before I can make a claim like that.

Appreciate the feedback here, but it's back work!

Best regards,
mfg

[late_actuary]

One simple tip for the P exam (I just started doing the practice problems just to see how I do "cold turkey"):

For problems in which you have to set up an equation and then solve for an unknown, it is sometimes less time consuming (and you are less likely to make an algebraic error) if you use the list of possible answers to simply "check" if it works with your equation as opposed to actually solving for the variable using algebra.

Case and point -- problem 4 in:

Questions: http://www.beanactuary.org/exams/preliminary/exams/syllabi/ExamPSamplequestions.pdf

Answers: http://www.beanactuary.org/exams/preliminary/exams/syllabi/ExamPSamplesolutions.pdf

If you look at the solutions, they show the more conventional technique for solving a problem like problem #4. A faster way is to stop once you obtain the equation, and just check which x value is the correct one. This is especially true for more complex equations.

[aaronhsmith]

One simple tip for the P exam (I just started doing the practice problems just to see how I do "cold turkey"):

For problems in which you have to set up an equation and then solve for an unknown, it is sometimes less time consuming (and you are less likely to make an algebraic error) if you use the list of possible answers to simply "check" if it works with your equation as opposed to actually solving for the variable using algebra.

Case and point -- problem 4 in:

Questions: http://www.beanactuary.org/exams/preliminary/exams/syllabi/ExamPSamplequestions.pdf

Answers: http://www.beanactuary.org/exams/preliminary/exams/syllabi/ExamPSamplesolutions.pdf

If you look at the solutions, they show the more conventional technique for solving a problem like problem #4. A faster way is to stop once you obtain the equation, and just check which x value is the correct one. This is especially true for more complex equations.

:iatp: