What's new

# Questions on memorizing t and z scores

#### Tinos

##### New Member
David and other fellow posters:

Do we get z-table and t-tables as an appendix on the exam, or do we have to memorize the critical values at certain confidence intervals?
(i know things like 1.645 come up a lot but I don't know the z-scores)

If we have to memorize, could you please again list the common stats for them?

Thanks!

-Chris
tinos88

#### David Harper CFA FRM

##### David Harper CFA FRM
Staff member
Subscriber
Hi Chris,

No, you don't need to memorize critical values (this policy will likely continue, you can ignore the mention of lognormal as that has been demoted in 2008). Most likely: you will be given the lookup value or, less likely perhaps a lookup table snippet. Yes, by sheer repeated exposure, -1.645 (1.654) & -2.33 should be second nature.

About the student's t table, I like to remind: if you look at it, as a practical matter, we often don't need it. At 90+% c.i. and sufficient d.f. (or more!), the "sweet spot" is around 2-3, so in *most* cases we recognize a high t (e.g., 8 or 15 or 25) as prima facie cause for rejecting the null. What i mean is, if the question says like, with 95% confidence and 20 d.f., t-stat returns, say, 6 or 8 or 25, we don't need a table. We can reject high values like this (i.e., reject the null that population mean is 0 or whatever we guessed for the population mean). Doubt creeps in at 3 or 2 or 1.5...

David

#### Tinos

##### New Member
OK, thanks....

This brings up probably a more crucial (and disturbing) point...I am having trouble determining *when* to use student-t methodology vs z-score vs chi-squared vs F-score, etc...is there a simple, easily memorized way for me to differentiate (i.e., when I see the that sample variance is given" I automatically know its z-score, so then I use this formula to get test statistic to solve problem, etc). I am trying to work through 2007 questions and am having a very tough time.

I'm thinking somewhere in the notes there probably is a table comparing all the different forms of test statistics with columns of "When to use", "How to perform", "Important Formulas for each", etc....something we could print off and attempt to commit to memory as we work through sample problems. Is this is the study guide or somewhere else and I just haven't gotten to it yet?

-thx..

#### David Harper CFA FRM

##### David Harper CFA FRM
Staff member
Subscriber
Hi Chris,

1. In the quant notes, pages 38 - 40+ (see TOC), I give one page summary for each; any shorter short-circuits an understanding...

2. Also, in the quant B movie (which has several slides devoted to this). For example below, the final slide is is a quick summary of purposes.

3. Finally, I really do encourage you read the "tips" newsletter that attaches to each screencast. The newsletters are no boilerplate, I write each freshly to highlight the most important ideas. I agree with you, this is "must know" so that's why i devoted a section in the tips letter on this. I think my summary here on the distributions is pretty succinct:

http://www.bionicturtle.com/learn/article/frm_episode_2_quant_b_econometrics/

(4. and finally, after all the notes are published, I will publish a master formula document that will contain these)

http://www.bionicturtle.com/images/forum/lastpage_quant_c.png

#### Tinos

##### New Member
This is perfect...thank you!

I will start making it a point to read the tips newsletters from now on...