Let me start off by saying that I AM NOT ASKING YOU TO DO MY HOMEWORK FOR ME. Too often when I post questions like this people don't answer the question and tell me to do my own homework. I am not asking you to do it for me, I am asking you to help show me how it is done because I don't understand. So here is the problem I am struggling with:

Three patients are enrolled in a clinical trial. Each patient either has a complete response to initial treatment or not. The probability of a complete response for any patient is 0.6 and the response of each patient is independent of the others. Let X denote the number of complete responders out of the first two patients enrolled. Let Y denote the number of complete responders out of the second and third patients enrolled.

a) Find the conditional distribution of Y given X=1.

b) Find cov(X,Y).

I know how to find a conditional distribution and covariance in a general case, but I don't know what to do in this case. How do I set this problem up? Is the joint distribution multinomial, like so:

$$\frac{3!}{x!y!} (0.6)^x(0.6)^y$$

Or is it something else? I honestly don't know. Anyone give me a pointer?


There are just $2^3 = 8$ possible outcomes here, where an outcome tells whether or not each of the three patients is a complete responder. You can list them all, compute their probabilities and the values of $X$ and $Y$, and calculate everything using the definitions.

  • $\begingroup$ I am having some trouble with this. First of all, how can there be 8 possible outcomes? So far as I can tell, there are only 7 possible outcomes. (X=0,Y=0), (X=1,Y=0), (X=0,Y=1),(X=1,Y=1),(X=2,Y=1),(X=1,Y=2), and (X=2,Y=2). What is the eighth outcome? $\endgroup$ – Abe Nov 3 '13 at 19:18
  • $\begingroup$ Ah, I see. There are two different ways for (X=1,Y=1) to happen. $\endgroup$ – Abe Nov 3 '13 at 19:29

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.