Confidence interval problem with unknown sample size information I am trying to answer a question about the following situation:

We have that a poll says that 36% of the people is going to vote for the candidate X. But there is a journalist that argues that 40% of the people are going to vote for the candidate X.

We have to give a confidence interval (95% of confidence) and say if the journalist could be right.
I don't have a clue how to get the confidence interval when you don't know the sample size.
[The correct answer for the interval is given as (0.3328, 0.3872).]
 A: You simply can't obtain the confidence interval in the answer from the information you provide. 
There are a variety of possible binomial proportion confidence intervals -- e.g see
https://en.wikipedia.org/wiki/Binomial_proportion_confidence_interval
-- from which you probably need the normal theory one.
... but whichever one you need, all of them depend on sample size.
The provided answer implies that the sample size was between 1192 and 1200 (assuming normal approx and no continuity correction) but if you don't have that information anywhere you had no way to get that interval.
Incidentally if I had a question like this in say an exam I would either


*

*assume a sample size (or even several different ones) in order to demonstrate I knew how to do it at some given sample size; or

*give the sample size that would make the difference (either the largest sample size consistent with the journalist's claim, or the smallest sample size inconsistent with the journalist's claim -- they should differ by only 1)
(or if time permitted, both of those things)
That way, even if you somehow missed the sample size detail somewhere, you may still get credit for showing you knew how to do it.
On the other hand, if I was doing a question like this for exam practice I'd just make up a sample size (typically in the ballpark of 1000) and do it.
