4

For a normal curve, how much of the area lies within 1.5 standard deviations of the mean? I already know about the 68–95–99.7 rule, and see that it should be between 68% and 95%. I also know that it should be closer to 95%, so I estimate it to be around 80%. Can someone give me a hint on how to find the answer?

KingLogic
  • 1,441
  • 2
    Look it up in a table or use a programming solution. For example this table gives $P(X \le 1.5) \approx 0.9332$ though you want $P(-1.5 \le X \le 1.5)$ – Henry Jun 22 '18 at 22:57

1 Answers1

2

You're close. It's about 87%. And see: probit.

Franklin Pezzuti Dyer
  • 39,754
  • 9
  • 73
  • 166
rulnick
  • 44
  • Combining henry's comment and rulnick's answer, I wanted to figure out how to get this answer using R, to compute $P(-1.5 < X < 1.5)$:

    > pnorm(1.5) - pnorm(-1.5) => [1] 0.8663856

    > 1 - 2 * pnorm(-1.5) => [1] 0.8663856

    This page has an interesting simplification, $1 - 2 * P(X < -1.5)$, as well.

     – dossy Oct 18 '21 at 15:03
  • Could you show your work or explain further, please? – 2540625 Sep 12 '22 at 20:29
  • For a random variable with density $f$, mean $\mu$, standard deviation $\sigma$, the area under $f$ within $k$ standard deviations of the mean is $\int_{\mu-k\sigma}^{\mu+k\sigma} f(u),du$, or $F(\mu+k\sigma)-F(\mu-k\sigma)$, where $F$ is its distribution. For a normal r.v., $F(u)=\Phi((u-\mu)/\sigma)$, where $\Phi$ is the standard normal distribution, so the area of interest is $\Phi(k)-\Phi(-k)$, or $1-2\Phi(-k)$ since, due to $f$'s symmetry, $\Phi(k)=1-\Phi(-k)$. Here, $k=1.5$. – rulnick Sep 14 '22 at 05:59