Sample size calculation for cross-sectional survey

I'm a grad student working on a project that has us send out surveys to doctors across a specific geographic region. I need to do some sample size and power calculations.

There are approximately 5,122 doctors in this area. I would like report alpha at 0.05 and 0.01 and power at 95% and 99 %.

The purpose of the study is to describe the clinical practice of doctors in this area based on the self-report surveys.

I am wondering if I can use stata for this calculation. I looked up the command sampsi but that is for comparing two groups. Is there a command in stata for this?



Ambassador to the humans
What is your response variable? When you say you want to achieve a certain power you also need to specify what the smallest difference is that you care about detecting.

Basically we need more information.
Thanks for your reply. This is a descriptive study so there isn't a response variable. I am wondering if I am making this more complex than it needs to be and confusing the terms power, alpha, margin of error and confidence intervals.

If I want to have a confidence level of 95% and a margin of error of 5%, can I use stata to calculate how many participants I need to send out surveys to?