stats breakdown

[Sox34]

Have any of the study manuals or anything made any predictions about possible # of ?s for the stats material by topic? I know that the syllabus percentage would suggest about 9 ?s or so, but wasn't sure what ppl are thinking in terms of some topics being tested more than others.

[Didon]

Have any of the study manuals or anything made any predictions about possible # of ?s for the stats material by topic? I know that the syllabus percentage would suggest about 9 ?s or so, but wasn't sure what ppl are thinking in terms of some topics being tested more than others.Sorry, I don't have an answer. I'm just bumping your thread hoping that someone else knows.
I'd be happy if people would just be consistent in their definitions of sample variance. :wall: I'd hate to miss a question just because I divided by n-1 instead of n.

[JohnK]

I'd be happy if people would just be consistent in their definitions of sample variance. :wall: I'd hate to miss a question just because I divided by n-1 instead of n.

Who defines sample variance as (sum of squared deviations) / n ?

As far as I know, if you're computing a sample variance you should always divide by n - 1. Maybe you are confusing sample variance with population variance in some cases?

John

[Sox34]

Who defines sample variance as (sum of squared deviations) / n ?

As far as I know, if you're computing a sample variance you should always divide by n - 1. Maybe you are confusing sample variance with population variance in some cases?

John

Actually, sample variance can be either /n or /(n-1) depending on whether or not you are dealing with a biased or unbiased estimator. Also, when assuming that the empirical distribution is the underlying distribution, the /n formula is called the sample variance.

[Sox34]

Ok, this is what i'm thinking in terms of the breakdown based solely on the percentages given on the syllabus...

8-10 Total ?s

2-4 ?s on Method of Moments, MLE, estimation criteria, etc
2-4 ?s on hypothesis testing
1-2 ?s on order stats from a sampling distr. like F-stat, T-stat,etc
1-2 ?s on linear regression

I know these percentages don't add up on the low and high end to 8-10 ?s...

what do you all think? does this seem right from the %ages given? Also, how accurate have the %ages been in the past in terms of questions asked?

[Sox34]

:bump: