Can you combine survey responses for a Confidence Interval

[lincoln5]

Hi, I'm doing basic analysis on a national survey but would like to combine responses to a question and was wondering if I can report a confidence interval for the combine value. Specifically, I gave many choices for when a speech pathologist begins treating a patient with neck cancer. I have percentages and confidence intervals for each response choice, "I begin 3 months before radiation therapy" and "1 month before", but I want to combine the two and just report how many begin before radiation, regardless of how long before. Can I do this, or does the fact that they selected specific choices not let me do confidence intervals on them. Any help would be appreciated. Thanks so much!

[signum]

Hi,

I hope I understood your description of the problem properly and if I did then I don't see reasons why you couldn't report the combined percentage. After all, you would then just be reporting the confidence interval of the "percentage of people who begin 1 months before or who begin 3 months before" and then you would call it something like "percentage of people who begin before", right?

Maybe it's just me but I can't see a problem here.

[lincoln5]

Hey signum,
Thanks so much! That's what I thought too but I just wanted to make sure that I could provide a confidence interval for that combined percentage. Thanks for your help!