I feel like I am spinning my wheels a little on the quantatitive methods sections. I am really trying to understand (and be able to perform) the processes to perform different statistical tests, but in the back of my mind I keep thinking “There is no way they are going to expect people to do this on the actual test.”
For example, I will try to get into how to calculate the degrees of freedom for t-tests versus Chi Squared versus Chi Squared comparing variances of two samples with unequal sample sizes.
Am I wasting my time here? It seems like all they may be getting here is for candidates to generally understand when to apply each test and what they can be used to test for, but that they wouldn’t go beyond a simple z-test or t-test on the actual exam. Am I being overly optimistic here?
Thanks. (BTW, this is not a question about what appears on the actual tests. Just a general question as to the level of specificity expected.)
Based on these LOSs, my understanding is that the Institute is more interested in your ability to formulate a hypothesis given a set of facts (this is critical, because everything that follows depends on it), then “identify” the appropriate test statistic and understand the decision rule /interpretation of results.
So even if you can’t memorize the exact long winded formulas, as long as you can identify the test statistic (I seriously doubt the answer choices will take one single formula and make trivial modifications to it and ask you which one is the correct one; ) and do the above you should be fine.
Realistically, the formulas will stick if you do some practice problems every so often. They’re not particularly bad, though. I absolutely think you should know how to conduct a test of hypothesis–formulating the correct null and alternative, to calculating a test statistic, and finally, making a conclusion with practical meaning (i.e. go beyond saying reject Ho). The practice material should have questions for you to do these sorts of tests.
By calculate degrees of freedom, do you mean calculate a test statistic? Determining degrees of freedom is pretty direct.
Consider the weighting of Quant relative to the other LOS’, do some practice problems, and move along. Doing nothing in Quant yields you an expected score of 33%. Putting in a ton of time might improve your Quant score, but with a high cost. Go study accounting, equity, derivatives, etc.
I agree not to spend too much time on something, especially if there are parts of the curriculum that have yet to be studied. However, a decent understanding of most of the statistics in QM at level 1 will be helpful in level 2. Additionally, the QM at L1 uses space to teach about DCF and other useful concepts that are pervasive in the rest of the curriculum.