Stan Brown <[EMAIL PROTECTED]> wrote in 
news:[EMAIL PROTECTED]:

> In article <[EMAIL PROTECTED]> in sci.stat.math, K L 
> <[EMAIL PROTECTED]> wrote:
>>I have a normal distribution with mean = 50 and standard deviation = 5
>>I want to divide this distribution into two separate normal distributions
>>each with mean = 25 so that when I add them I will get my original
>>distribution.
> 
> This makes no sense. When you combine two distributions, the mean of 
> the combined distribution must be between the means of the separate 
> distributions:
>      MU = (MU1*N1 + MU2*N2) / (N1 + N2)
> Therefore if the mean of the big distribution is 50, it _can't_ be 
> split into two distributions whose means are both 25.

That's what I thought he was talking about at first, but after reading the 
original post a couple times it looks like he's saying that he's got an RV, 
X~N(50,5) and he wants to write X as Y1+Y2, where Y1~N(mu1,s1) and 
Y2~N(mu2,s2).  In other words, "divide" here doesn't mean "partition."
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to