a response to a survey question is either yes or no. The responses are divided into ten categories, based on other information. So in category 1, there's 341 yes and 1023 no, cat 2 128 yes and 744 no, etc.
I would like to calculate some sort of confidence in the proportion of yes/no over the categories, like a confidence interval saying that for category 1, we are 95% sure the true value falls within this range. Do I perform either a Chi Squared test or a Simultaneous Multinomial Confidence Interval?
