Bournelli, equation for variance? (with picture!)
Maths and statistics discussion, revision, exam and homework help.
-
- Reputation:
- Thread Starter
- Peer Of The TSR Realm
- Location: London Posts: 10,406
Bournelli, equation for variance? (with picture!)So I was watching khan academy videos on Stats, and I noticed this:
why he does not calculate the variance of the sample he has first, and then approximate the variance of the normal distribution using the result?
Rest of the question is on picture in white text, any help much appreciated
Last edited by Moa; 17-05-2012 at 23:01. -
- Reputation:
- TSR Idol
- Location: London
Re: Bournelli, equation for variance? (with picture!)What is M_1 (or mu_1) at the very start? And what is he trying to show, the normal distribution approximation for the sum of 1000 choices of 0 or 1 where the probability of choosing 1 is 0.642? In any case, that's a freaking tiny variance that can't be right for whatever
-
- Reputation:
- Thread Starter
- Peer Of The TSR Realm
- Location: London Posts: 10,406
Re: Bournelli, equation for variance? (with picture!)mu_1 is mean of sample number 1(Original post by Hopple)
What is M_1 (or mu_1) at the very start? And what is he trying to show, the normal distribution approximation for the sum of 1000 choices of 0 or 1 where the probability of choosing 1 is 0.642? In any case, that's a freaking tiny variance that can't be right for whatever
He tries to show how many people have voted for a candidate (choice 1) and how many didn't (choice 0), by using this sample he then creates a normal distribution of all the sample means. -
- Reputation:
- TSR Idol
- Location: London
Re: Bournelli, equation for variance? (with picture!)I'm still not sure what he's trying to show, that final super tiny variance makes no sense to me, sorry. Was there anything prior to this question he built upon?(Original post by Moa)
mu_1 is mean of sample number 1
He tries to show how many people have voted for a candidate (choice 1) and how many didn't (choice 0), by using this sample he then creates a normal distribution of all the sample means. -
- Reputation:
- Thread Starter
- Peer Of The TSR Realm
- Location: London Posts: 10,406
Re: Bournelli, equation for variance? (with picture!)Maybe the video can help you, if you have time to skim through it:(Original post by Hopple)
I'm still not sure what he's trying to show, that final super tiny variance makes no sense to me, sorry. Was there anything prior to this question he built upon?
http://www.youtube.com/watch?v=a1Ye5...28115D3D8A2566 -
- Reputation:
- TSR Idol
- Location: London
Re: Bournelli, equation for variance? (with picture!)Ah, I see. It builds on earlier videos (or so he says), but he's working out how the sample mean varies. He has worked out the variance of the first sample, but by a 'stock' formula (Bin(n,p) has variance np(1-p)), which gives the same answer as your more general formula. I'm assuming you're okay with the rest, just the calculation of the variance was different to what you expected?(Original post by Moa)
Maybe the video can help you, if you have time to skim through it:
http://www.youtube.com/watch?v=a1Ye5...28115D3D8A2566 -
- Reputation:
- Thread Starter
- Peer Of The TSR Realm
- Location: London Posts: 10,406
Re: Bournelli, equation for variance? (with picture!)Yes that is it, its all clear now, I can see that the results of "mine" and his formulas are very close.(Original post by Hopple)
Ah, I see. It builds on earlier videos (or so he says), but he's working out how the sample mean varies. He has worked out the variance of the first sample, but by a 'stock' formula (Bin(n,p) has variance np(1-p)), which gives the same answer as your more general formula. I'm assuming you're okay with the rest, just the calculation of the variance was different to what you expected?
