Longitudinal Statistics


New Member
I'm performing a longitudinal study on allergen challenged (daily) mice (3 total) where I sample blood from the same mouse daily, over a week's time, and analyze them for leukocyte counts in an effort to show an increase in leukocyte count relative to number of exposures. These mice are not compared to a control group of non-challenged mice.

I've been under the impression that performing such a study (longitudinal) has enhanced statistical power over one where, say, three mice are sacrificed for each day, rather than repeatedly sampling from the same mouse. If this assumption is correct, under what statistical calculations am I expected to find this enhanced power? Moreover, being that improved standard error/deviation are the values i'm really after, is there a direct calculation of standard deviation or standard error where, perhaps, the n values can be increased, on behalf of it being a longitudinal analysis, and thus provide enhanced statistical power?