mcarthur724
Member
In going through the Gujarati readings I have noticed that many calculations require looking up numbers from the tables in the back of the book. Do they provide these statistics tables at the exam?
Are tables from back of statistics book provided at exam
- Thread starter mcarthur724
- Start date
cpaguy7,
Yes, if needed, GARP will provide because they cannot expect you to know critical lookup values. However, you should know:
one-tailed Pr(z>) = 5% @ 1.645
one-tailed Pr(z>) = 1% @ 2.33; i.e., the two VaR confidences that you can expect
two-tailed Pr(z>) = 5% @ 1.96 = one-tailed at 2.5%
two-tailed Pr(z>) = 1% @ 2.58 = one tailed at 0.5%; i.e., the two confidence intervals often used for test of regression coefficient
and, for critical t values, know that computed t's >> 2 or 3 will generally find significance (reject null) given sufficient d.f. and confidence; e.g., critical t = 5 or 8 can be eyeball rejected if decent sample and confidence 95/99
.... under stress, it's easy to forget that sometimes you don't need a lookup table. see this sample question @ http://forum.bionicturtle.com/viewthread/2528/ where UBS observed 29 exceedences against a mean (expected) of 2.5; we can compute the critical value, etc, but we don't need to, it's going to be large ... the critical value is just the number of (standardized) standard deviations from mean: if it's large enough, you are in the reject region, we can just "eyeball" reject
David
Yes, if needed, GARP will provide because they cannot expect you to know critical lookup values. However, you should know:
one-tailed Pr(z>) = 5% @ 1.645
one-tailed Pr(z>) = 1% @ 2.33; i.e., the two VaR confidences that you can expect
two-tailed Pr(z>) = 5% @ 1.96 = one-tailed at 2.5%
two-tailed Pr(z>) = 1% @ 2.58 = one tailed at 0.5%; i.e., the two confidence intervals often used for test of regression coefficient
and, for critical t values, know that computed t's >> 2 or 3 will generally find significance (reject null) given sufficient d.f. and confidence; e.g., critical t = 5 or 8 can be eyeball rejected if decent sample and confidence 95/99
.... under stress, it's easy to forget that sometimes you don't need a lookup table. see this sample question @ http://forum.bionicturtle.com/viewthread/2528/ where UBS observed 29 exceedences against a mean (expected) of 2.5; we can compute the critical value, etc, but we don't need to, it's going to be large ... the critical value is just the number of (standardized) standard deviations from mean: if it's large enough, you are in the reject region, we can just "eyeball" reject
David
mcarthur724
Member
This is helpful to know for my preparation. Thanks!
Similar threads
