Dear List Subscribers,
A researcher client is conducting a repeated measures experiment
on mice that are suffering from degenerative brain shrinkage. The client
starts with 10 mice at the beginning of the trial,but some of the mice die
at successive time points (measured in weeks) of the design. The client
wishes to compare a control group of diseased mice with a group of mice
that have been treated with a drug to slow the brain shrinkage. Her question
is what is the sample size necessary for the repeated measures experiment
(initial number of control and treated mice) to determine the efficacy of the
drug treatment with p = 0.05. The design is not necessarily balanced because
some of the mice die during the course of the trial, so, for example, 10 control
mice at the beginning of the trial may be 8 control mice at the end. She also
mentioned power at 80% or 95%.
Any advice would be greatly appreciated,
Eric
|