Typo in the Gujarati book

[blueturtle]

Hi all,

In my edition of Gujarati's Essentials of econometrics, the sense of Chebyshev inequality (eq. 3.20, page 57, and example 3.21) is wrong.

No big deal but I've already scratched my head in the past for such small details...

The official site of the book does not provide any errata list.

So if you spot any other errors, you may use this thread to report them.

 

[David Harper CFA FRM]

blueturtle,

(3rd Edition - right?) Can you clarify, Ex 3.5 looks correct to me? 2 s.d. implies 75%/25%. Thanks, David

 

[blueturtle]

The inequality 3.20 should be :

P(abs(X-mu_x) <= c.sigma_x) >= 1-1/c^2

instead of :

P(abs(X-mu_x) >= c.sigma_x) >= 1-1/c^2

 

[blueturtle]

Oops, I meant :

'The inequality 3.20 should be :

P(abs(X-mu_x) <= c.sigma_x) >= 1-1/c^2

instead of :

P(abs(X-mu_x) <= c.sigma_x) <= 1-1/c^2 '

 

[David Harper CFA FRM]

Ah, nice catch! I missed it, thank you (fortunately, my notes/slide are correct b/c i carried forward last year's usage which instead expresses the P[outside the region] which is a less than). David

 

[Atin]

Just noticed that the book does have an errata - of course, it does not have the one listed above, though.

http://highered.mcgraw-hill.com/sites/0072970928/student_view0/textbook_errata.html

 

[saket03]

Think another very small error is in Chapter 3 on page no. 67. Leptpokurtic has been defined as having slim or long-tailed and Platykurtic is defined as having fat or short-tailed. This shd actually be reverse.

 

[David Harper CFA FRM]

That's interesting. Please note in the FRM, leptokurtosis implies kurtosis > 3 (i.e., excess kurtosis > 0). Without disagreeing with you, just for exam purposes, i would note that leptokurtosis associates with "fat" or "heavy" tails.

But I find your note interesting b/c kurtosis is fourth moment function and really, I think, is a matter of the PEAKEDNESS of the function. So, leptokurtosis implies, first, higher peaks. Then, we typically say, okay, higher peaks implies fatter tails. But I *think* (do correct me if you know better) we really mean: longer, skinnier tails. It seems to me that, strictly speaking, fat tails is ambiguous; fat tails suggests a vertical metric in the tail but we really mean more density in the extremes (which involves a horizontal perspective) but this can look long and skinny (!?)

All in, as i think about this, "heavy tails" now seems more descriptive as it avoids the vertical fat/skinny idea. Curious if you would agree: leptokurtosis = "heavy tails"?

David

 

[saket03]

Thanks for this David.

Agree with you on the interpretation of "fat" tails as "heavy" tails.

I am very new to this Forum and this website but I have found it to be of immese value.

My Compliments to you!