Estimation of the power of the Wald test in order to compare the means of the l > atent trait in two groups of individuals
raschpower [, n0(#) n1(#) gamma(#) var(#) d(vector)]
Description
raschpower allows estimating the power of the Wald test comparing the means of two groups of patients in the context of the Rasch model. The estimation is based on the estimation of the variance of the difference of the means based on the Cramer-Rao bound.
Options
n0 and n1 indicates the numbers of individuals in the two groups (100 by default).
gamma indicates the group effect (difference between the two means) [0.5 by default].
var indicates the value of the variance of the latent trait [1 by default].
d is a vector containing the difficulties of the items [one row per item, one column - (-1,-0.5,0,0.5,1)' by default].
Example
. raschpower
. raschpower, n0(200) n1(200) gamma(0.4) var(1.3)
. matrix diff=(-1.47\-0.97\-.23\-0.12\0.02\0.1) . raschpower, n0(127) n1(134) gamma(0.23) d(diff) var(2.58)
References
Hardouin J.B., Amri S., Feddag M., Sébille V. (2011) Towards Power And Sample Size Calculations For The Comparison Of Two Groups Of Patients With Item Response Theory Models. Accept to appear in Statistics in Medicine.
Author
Jean-Benoit Hardouin, PhD, assistant professor EA4275 "Biostatistics, Clinical Research and Subjective Measures in Health Sciences" University of Nantes - Faculty of Pharmaceutical Sciences 1, rue Gaston Veil - BP 53508 44035 Nantes Cedex 1 - FRANCE Email: jean-benoit.hardouin@univ-nantes.fr