What is the correct statistical test to use for comparing proportions?

I am working on doing a test at work to see what the difference between the turnover rates of employees that interned and those that did not. We don't have a lot of great data so I'm using all the available data as a sample. I did a one tailed proportion test to see if the difference was significant, but I had a colleague tell me that this wasn’t reasonable because there was no sample error, and thus I couldn’t develop of confidence Interval. This confused me as you wouldn’t have a CI for that as the result is just the p value. He recommended I find a different test such as the chi square test so that I can get a sample error. Is this the right approach, any thoughts would help!