Stats Question/ Help? :)

[sprinkles]

Hello all,

 

So this is random but I have a stats exam tomorrow and I have a question and I dont know where else to post? lol

 

 

So my prof mentioned last minute that we need to memorize all info for 4 confidence levels. I have the 95% level memorized because thats what we have used all term but she said you should memorize the rest.

 

Since I already have so much to memorize, and since I dont know if it will even come up on the exam, I was wondering if you guys know of a way to 'calculate' or 'look it up' during the exam should it come up? My prof mentioned vaguely that its possible once but she never expanded.

 

So what I need to know is:

 

 

CL | Alpha | Z score

90% 0.10 +/- 1.65

95% 0.05 +/- 1.96

99% 0.01 +/- 2.58

99.9% 0.001 +/- 3.29

 

 

I know all the CL and all the alphas and the z-score for 95%. All I need to know is the z-scores for 90%, 99%, and 99.9%

 

 

Any tips/ideas?

 

Mucho appreciated!

[sprinkles]

nobody?

[soporific]

I haven't taken stats yet, but your best bet is to just straight up memorize it. You can think of a mnemonic just in case you doubt yourself.

 

For example, the z scores always increase along with the CL. Did you notice that the difference between the z score for 99/95 is .62, which is double the difference (.31) between 95/90? Since you know the z score for 95 you can calculate the ones for 90 and 99. And for 1.65, 1+5=6? Then if you add the z values for 90/99 you'll get 3.2_, which are the first digits of the z score for 99.9 . Since there are so many 9's you can remember that it ends in a 9 instead of a 3.

 

Or maybe I'm just crazy...

[osteon]

You can calculate it, but you then need to know a few things:

 

1) The general equation for a normal distribution (the equation for the curve itself)

2) Calculus

 

Basically you set the integral of the equation to whatever percentile you want and then solve for X to get your z score.

 

I suggest you just memorize it.