Standard Deviation

[PooyaH]

---

Members do not see this ad.

On destroyer QR #152:

After finding the mean and deviation (15.14285714), don't we have to first divide it by hte total number of values -1 (N-1) and then get the square root?
That's how I remember standard deviation from 2 years ago and I believe that's how we did it lol

Let me know what you think.

Thanks!

 

[joonkimdds]

I believe Destroyer is wrong.
Destroyer makes many mistakes in math.

 

[PooyaH]

OMG I posted this like a year ago! LOL

 

[Shinpe]

Standard deviation is the square root[ sum { (each term - mean)^2} /(N-1)]. As should be on wikipedia. The reason you use N-1 is as a safety factor to conservatively calculate your standard deviation (have a larger standard deviation than if you had just calculated the mean with N). I have seen people using N here and there but the most commonly accepted way of doing it is using N-1. There is no other mathematical reason for it.

 

[pbure9]

Standard deviation is the square root[ sum { (each term - mean)^2} /(N-1)]. As should be on wikipedia. The reason you use N-1 is as a safety factor to conservatively calculate your standard deviation (have a larger standard deviation than if you had just calculated the mean with N). I have seen people using N here and there but the most commonly accepted way of doing it is using N-1. There is no other mathematical reason for it.

I was just going to make a post about this. So I guess we should be using N-1, not N?

 

[mmorris86]

So wikipedia says its N, is that the final answer?

 

[pbure9]

lol. I don't know. I did math destroyer problems doing N to get the right answer, used in on Acheiver and they wanted N-1. I guess they'll have either answer. It's just a pain if you have to do it twice.

 

[Shinpe]

Oh I guess I never actually checked wikipedia but figured they'd have the more widespread definition. I know that in biostatistics books (at least mine, glantz) it's N-1. Also, microsoft excel uses N-1.

And I trust those more than wikipedia. I just hope that people who write actual DAT questions use the right equation (which I think is N-1), not the other one. They probably do.

 

[tsitneD]

actually there is a difference

Population standard deviation. The population standard deviation (often represented by the Greek letter sigma) is measures the variability of data in a population. It is usually an unknown constant. The formula is:

​

Sample standard deviation. The sample standard deviation (usually represented by S) measures the variability of data in a sample. It is easy to compute (compared to a population standard deviation) because it is based on a small and manageable sample. The formula is:

​

 

[tsitneD]

so i think on DAT we use sample deviation because we have a given set of exact number to calculate without any flaws or uncertainty.

 

[mmorris86]

so i think on DAT we use sample deviation because we have a given set of exact number to calculate without any flaws or uncertainty.

Nice! Thank you, this makes sense

 

[pbure9]

It is easy to compute (compared to a population standard deviation) because it is based on a small and manageable sample.

that's what i needed to see. Sounds good. 👍

Thanks guys