Sampsize is a tool that computes sample size and power for surveys, comparative studies including non-inferiority trials.
Sampsize computes sample size for prevalence surveys with finite population correction, sample size to detect a minimum number of events, sample size and power for one-sample and two-sample comparison of percentages and means, corrections for cluster sampling designs, sample size and power for unmatched and 1:1 case-control studies, sample size and power for non-inferiority trials.
This is a fast command line programme. Sampsize is written in C and depends only on libc. All computations are in double precision. A certification script is included. Sampsize is released under the GNU General Public Licence.
Here are some key features of "Sampsize":
· Sample size for prevalence surveys
· Finite population correction
· Sample size to detect a minimum number of events
· Exact binomial confidence intervals
· Sample size and power for one-sample comparison of percentage to hypothesized value
· Sample size and power for two-sample comparison of percentages
· Sample size and power for one-sample comparison of mean to hypothesized value
· Sample size and power for two-sample comparison of means
· Correction for cluster sampling designs
· Sample size and power for unmatched and 1:1 matched case-control studies
· Minimum/Maximum detectable odds-ratio in case-control studies
· Optional Continuity Correction
· Sample size of non-inferiority trials with comparison of means and percentages
What's New in This Release:
· Continuity correction option for comparing two percentages (sample size and power). Default is no correction.
· Implemented sample size and power for non-inferiority (equivalence) tests of two percentages and two means.