3
$\begingroup$

In a certain population, 18% of the people have Rh-negative blood. A blood bank serving this population receives 95 blood donors on a particular day.

So my teacher posted a solution to a problem below which solves the probability of getting 10 or less donors that are Rh-negative. I was wondering if I wanted to instead find the probability of 10 or MORE donors that are Rh-negative, how would I do so?

This was my attempt from following the teacher's logic

p(x≥10) = p(x>9.5) = 1 - p(Z>((9.5-17.1)/3.745)) = 1 - (Z>-2.03) = 1 - 0.2118 = 0.7882

Can someone confirm if I did this right, and if not, teach me how? I'm still a little lost.enter image description here

$\endgroup$
2
  • 1
    $\begingroup$ First, pay attention that 18% $\ne$ 15%, and 95 $\ne$ 100. $\endgroup$ – user158565 Jul 29 '19 at 1:18
  • 1
    $\begingroup$ p(x≥10) = p(x>9.5) $\approx$ p(Z>((9.5-17.1)/3.745)) = p(Z>-2.03) =0.9788 $\endgroup$ – user158565 Jul 29 '19 at 3:20
2
$\begingroup$

Let $X \sim \mathsf{Binom}(n = 95, p = .18).$ Then you seek $P(X \ge 10) = 1 - P(X \le 9) = 0.9841.$

Directly in R, where pbinom is a binomial CDF, we have

1 - pbinom(9, 95, .18)
[1] 0.9840775

Using a normal approximation with continuity correction, you should begin with $P(X \ge 10) = P(X > 9.5) = \cdots .$ You have the right idea, except I guess you got mixed up using normal tables. In particular, $P(Z>-2.03) = 0.9788 \ne 0.2218.$

1 - pnorm(-2.03)
[1] 0.9788217

Below is a plot of the relevant part of the PDF of $\mathsf{Binom}(95,0.18)$ along with the normal distribution that matches the binomial mean and standard deviation.

enter image description here

x = 0:35;  PDF = dbinom(x, 95, .18)
plot(x, PDF, type = "h", lwd=2)
  abline(h=0, col="green2")
  abline(v = 9.5, col="red", lty="dotted")
mu = 95*.18;  sg = sqrt(mu*(1-.18))
  curve(dnorm(x, mu, sg), add=T, lwd=2, col="blue")
$\endgroup$

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.