Calculating SD/Variance

[MedHopeful234]

Hey Guys,

with regards to calculating Variance/SD do you use n or n-1?

View attachment 25543

I recall Chad mentioned to do n-1 and a problem in achiever also used n-1, but then in other sources (Destroyer I believe) it was done with n.

WOW SO SORRY FOR POSTING THIS MULTIPLE TIMES! MY INTERNET KEPT DISCONNECTING and I couldn't tell if I had posted the thread and ended up doing multiple posts!

PLEASE DELETE REPEATED POSTS MODS!

---

Members don't see this ad.

 

[Conflagration]

Hello! I'm a math major.

You use n-1 for calculating variances and standard deviations.

When you do a summation set of differences of means, and divide by n, you find the average of the differences between the means.

On a more practical level, if you have 5 objects, for example, you would have 4 spaces in between those 5 objects.

Subtracting n by 1 removes some of the bias(recall that a mean is heavily swayed by outliers), making it a good estimator for the population variance and standard deviation.

Best of luck to you!

 

[ToBeDDSS]

From what i remember, you use n-1 if you have a sample and use n if you have a population. In all stats book they usually specify if it were a sample or a population.
hope this helps

 

[datthirty]

From what i remember, you use n-1 if you have a sample and use n if you have a population. In all stats book they usually specify if it were a sample or a population.
hope this helps

That is what I learned in college.