Confidence Interval help please - mean or proportion?
Firstly apologise for the length, any help would be greatly appreciated though.
I am interested in the average proprtion of debt that is paid back to me.
I have a population of people who have been in debt to me. From this population i took a random sample of people and recorded the proportion of debt which had been paid back.
Hypothetically, I have a population of 5000 and a sample of 200. The sample suggests that the average amount of debt paid back is 70%.
Should I use proportions to construct the confidence interval (i am not sure if proportions should only be used when the result is success or failure) or should i use means and the mean here would be 0.7?
Also, the vast majority of cases in my sample are 0% or 100%, does this mean that my sample is not normally distributed and does it affect my confidence interval.
Thank you very much in advance