0
$\begingroup$

Here is the question I need help on :

Let $X$ be a binomial random variable with p = 0.5 and n = 100.

Give $P(X \geq 60)$ rounded to two decimal places without using a calculator (by using approximations I assume).

I easily find that $ \displaystyle P(X \geq 60) = 1 - \sum_{k=0}^{59} \binom {100} {k}(0.5)^{100}$ or alternatively $\displaystyle \sum_{k=60}^{100} \binom {100}{k} (0.5)^{100}$. But I can't seem to find a good approximation of this that doesn't require the use of a calculator or a little program.

Thanks for the help !

$\endgroup$
  • 1
    $\begingroup$ I would guess they are expecting you to use a normal approximation. $\endgroup$ – André Nicolas Nov 6 '14 at 17:37
  • $\begingroup$ That leads to be able to approximate the cumulative distribution function of a normal variable, any clue? $\endgroup$ – Raven Nov 6 '14 at 17:41
  • $\begingroup$ you can use the z-score and then use the tabulated values to generate the required probability. $\endgroup$ – Chinny84 Nov 6 '14 at 17:43
1
$\begingroup$

You can make an approximation by the normal distribution:

$\large{1-P(X\leq 59)\approx 1-\Phi \left( \frac{59+0.5-50}{\sqrt{100\cdot 0.5 \cdot 0.5}} \right)}$

$\Phi(z)$ is the cummulative function of the standard normal distribution.

0.5 is the continuous correction factor.

Formula for approximation the cdf of the standard normal distribution:

$\large{P(X\leq x) \approx 0.5 \cdot \Bigg( 1 + \sqrt{1-e^{- (\sqrt{\frac{\pi}{8}} \cdot x^2)}} \Bigg)}$

enter image description here

Normalverteilung (german)=normal distribution

$\endgroup$
  • $\begingroup$ Thanks. I came up with this but I couldn't approximate $\Phi(9.5/5)$ without a table, is there a simple way to do it ? Or can I approximate it by $\Phi(2)$ (I mean, am I supposed to know $\Phi(2)$ by heart ?) $\endgroup$ – Raven Nov 6 '14 at 18:01
  • $\begingroup$ Nevermind, $\Phi(1.96)$ = 0.975. Thanks a lot. $\endgroup$ – Raven Nov 6 '14 at 18:15
  • $\begingroup$ Just wait. In general you need a table. But long ago I have found a formula, which approximate the standard normal distribution. The picture visualizes it. It is your decision, if it is exactly enough for you. I have made an edit. $\endgroup$ – callculus Nov 6 '14 at 18:29
  • $\begingroup$ I've seen the formula, thanks that's a great visualisation. But, without a calculator, it's impossible to calculate the value of this $\endgroup$ – Raven Nov 6 '14 at 19:15
  • $\begingroup$ Yes, without a calculator you are lost. You can learn the standard values of z (1.96, 2.33, ...) by heart and hope that this what you need. But this is not really a sure road. The posted formula is the best simplification I know. $\endgroup$ – callculus Nov 6 '14 at 19:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.