We use the program MARK to estimate population size and survival of lizards. The program output is an estimate of various parameters and their SE. For example, the survival on a given year in a specific location. We would like to compare and decide whether it is significantly different from an estimate of the same parameter in a different year or a different location (these estimates are also provided in the output). I thought to calculate Confidence intervals (CI) around the means and see if they overlap, to tell if they are different. The MARK program calculates these parameters and the SE without providing us with a sample size. So because I don't have a sample size associated with the SE, I cannot obtain the standard deviation, that is normally used to calculate the CI. Is there any solution around this problem?
Thanks, Amos
Sorry - The answer is very simple - Mark provides in the right hand side of the output the Upper and Lower Confidence Interval, so no need to calculate it...
Amos
Standard deviation is fixed, but estimated. It is with regard to a population. Standard error comes from standard deviation and sample size, and it changes, becoming smaller with increased sample size. It would seem that you would have entered the sample data and know the sample sizes. Curious.
Anyway, if you are looking at survival rates, I guess the populations are very skewed, and you'd want to take that into consideration.
Enjoy your project. Cheers.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size