| swGlmSiz | R Documentation |
swGlmSiz is a wrapper function for swGlmPwr. For a given design, number of clusters, power and fixed and random effect parameters, swGlmSiz computes the number of individuals per cluster needed to achieve the desired power.
swGlmSiz(power, alpha=0.05, ...)
power |
numeric (scalar) Desired power. Must be between |
alpha |
numeric (scalar): Two-sided statistical significance level. |
... |
Arguments to be passed to |
swGlmSiz calls swGlmPwr iteratively to find the number of participants per cluster-period needed to achieve power power. Note that all warnings and messages normally generated by swGlmPwr are turned off. Also, the argument retDATA is set to FALSE. It is recommended that after using swGlmSiz, the user call swGlmPwr with the recommended sample size to view any warning messages and/or to get the data normally returned when retDATA is TRUE.
An error in swGlmPwr causes swGlmSiz to terminate and the error message from swGlmPwr is printed.
numeric (list): swGlmSiz returns a list of the following:
n |
scalar: number of individuals per cluster-period required to achieve the desired power. |
power |
scalar: exact power, as computed by |
Avi Kenny, James P Hughes
swGlmSiz(power = 0.8, design = swDsn(clusters=c(4,4,4,4)),distn = "binomial",
fixed.intercept = log(0.1/0.9), fixed.treatment.effect = log(0.9),
fixed.time.effect=0, tau = 0.2)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.