Normal distribution - how to solve P(-b<X<b)=0.95

1.2k Views Asked by At

$X\sim N(2,3^2)$

How do you find $b$ where $P(-b<X<b)=0.95$ other than trial and error? You can't directly transform to $z$ because if you find an appropriate $z$, transforming back will give you the difference between $b$ and $z$, not $\pm z$.

I know it's a simple question but I'm stuck in a rut.

1

There are 1 best solutions below

3
On BEST ANSWER

I don't think there's a purely analytical way to do this, but perhaps I can offer something a little more systematic than guess and check:

Lets first consider the much simpler problem of a mean $0$ normal RV with variance $9$.

In that case, we know from lookup tables that $b\approx 1.96\times 3$ will solve our problem. However, the shift upwards by two reduces the probability to:

$$P(-b<X<b)=\Phi\left(\frac{b-2}{3}\right)- \Phi\left(\frac{-b-2}{3}\right)<0.95$$

Now, lets augment this $b$ by another value $\delta>0$ so that:

$$P(-b<X<b)=\Phi\left(\frac{b-2+\delta}{3}\right)- \Phi\left(\frac{-b-2-\delta}{3}\right)=0.95$$

Therefore, you can simply solve (numerically), the equation above (I did it using Excel's Goal Seek) as a function of $\delta$ to get 0.95. I got 6.967891559 from Excel, similar to you. Unfortunately, the Gaussian CDF is not an easy function to work with analytically.