• We are pleased to announce that the winner of our Feedback Prize Draw for the Winter 2024-25 session and winning £150 of gift vouchers is Zhao Liang Tay. Congratulations to Zhao Liang. If you fancy winning £150 worth of gift vouchers (from a major UK store) for the Summer 2025 exam sitting for just a few minutes of your time throughout the session, please see our website at https://www.acted.co.uk/further-info.html?pat=feedback#feedback-prize for more information on how you can make sure your name is included in the draw at the end of the session.
  • Please be advised that the SP1, SP5 and SP7 X1 deadline is the 14th July and not the 17th June as first stated. Please accept out apologies for any confusion caused.

Confidence Intervals- Two population proportions

N

Newcomer101

Member
I am having difficulty deriving the pivotal quantity for the above confidence interval. (pg 24 chapter 11)

Taking two random samples:
A~bin(n1,p1) approximates to N(n1p1,n1p1q1)
B~bin(n2,p2) approximates to N(n2p2,n2p2q2)

Then A-B approximates to N(n1p1-n2p2,n1p1q1+n2p2q2)

Therefore, should it not follow the pivotal quantity should be:
(n1p1-n2p2)-(theta1-theta2)
sqrt(n1p1q1+n2p2q2)

Any help most appreciated.
 
I think you might be getting confused between the distribution of X and the distribution of p(hat).

You have defined the distribution of A and B, which are equivalent to X1 and X2. These are binomial random variables with expected values n1p1 and n2p2, and variances n1p1q1 and n2p2q2. (You've used p where the Core Reading uses theta.)

Using a normal approximation, if you standardise the distribution of X1-X2 (to get the number of standard deviations X1-X2 is from its mean), you'd get:

[(X1-X2)-(n1p1-n2p2)]/sq rt(n1p1q1+n2p2q2) is approx N(0,1)

We can't use this to get a confidence interval for p1-p2 directly, so we need to use the distribution of the difference between the sample proportions p1(hat) - p2(hat), (where p1(hat) =X1/n1).

The expected value of p1(hat) - p2(hat) = p1-p2 and the variance of p1(hat) - p2(hat) = p1q1/n1 + p2q2/n2.

Standardising, we get:

[( p1(hat) - p2(hat)) - (p1-p2)]/sq rt (p1q1/n1 + p2q2/n2) is approx N(0,1)

If we want to find a 95% confidence interval for p1-p2, we use:

p1(hat) - p2(hat) +/- 1.96 sq rt (p1(hat)q1(hat)/n1 + p2(hat)q2(hat)/n2

(We have to use the sample proportions to estimate the standard deviation.)
 
Back
Top