0

I have population data (not sample data) and I'm looking at incidence over time in the population. I have small n's (for the incidence and for the overall population). For example, I have an incidence of 9, occurring in a population of about 40,000. Because I am dealing with small n's, would it be appropriate to use confidence intervals and trend analysis, even though I am not actually working with a sample?

Kelly
  • 1
  • 2
    Are your study conclusions going to be limited to what happened *just to this population* and *only for this time interval*? If not, then you have a sample. Regardless, how can we advise you about what procedures to use? You haven't told us what you're trying to learn. – whuber Apr 27 '16 at 13:48

1 Answers1

1

If you are really dealing with population data than there is nothing to estimate, since there is no uncertanity. The event happened in the population exactly 9 times.

Check also: Statistical inference when the sample "is" the population and What is the difference between a population and a sample? (as @whuber may be right that you may be misinterpreting your sample as population).

Community
  • 1
Tim
  • 108,699
  • 20
  • 212
  • 390