Stats Tables

Hi, In the AnalystNotes mock exam, I came accross a question that (I think) required looking up in one of the stats tables in the back of the quant book. It had to do with whether some correlation number was significant for some value of alpha. Am I missing a way to solve this w/o looking at the tables? If not, are questions like that possible on the actual exam, and if so - are the stats tables provided? Thanks, Ed

if you nned any numbers from tables they will provide you with a small excerpt of info (the correct relevent values and some wrong info). They don’t provide the whole tables.

Ed, please attempt to make full use of the CFAI website, including its FAQ. I recommend knowing the z-scores for 90%, 95% and 99% Consistent with what Super said: http://www.cfainstitute.com/cfaprog/faq/faqs_candidates.html##89 In quantitative methods we often refer to the z-table, t-table, and F-table. Will these tables be provided during the exam? No. CFA Institute will not provide distribution tables with the exam. If a question requires information from a specific distribution table, that information will be provided with the question, but the tables themselves will not be provided. Similarly, time value of money tables (present value, future value, etc.) are neither needed nor provided. The two approved calculators are capable of performing any required present value and future value calculations.

They could approve a calculator that would give you distribution quantiles. It’s a little unfortunate that my cell phone is wildly more powerful than the calculators that they will let you bring into the exam. Why do they feel such a need to hobble people in this unrealistic way? It would be a much more realistic test if everyone got to use Excel.

To take what hiredguns said a little further, as a shortcut on the exam, as long as n > 30, you can generally use the z-scores in place of the t’s. As long as they’re not too close, this saves a lookup (and a buit of time). For example, if you had a t of 3.2, you wouldn’t need to look it up to see if it was significant - it will be. One N gets large, the distribution converge to the z.