Dear All,
Sorry to trouble. I have measured the same variable in subjects at 5
(equally spaced) time points. There are two treatment parameters
(stimulus/no stimulus and the other no inhibitor/inhibitor A/inhibitor B)
giving a total of six possible treatment groups. Treatments were applied
immediately after the first observation. Graphically (and logically) the
variable increases over time in some groups. There are some missing
observations. I want to describe specific and groups and then make
comparison between specific groups. Mistakenly in the past I have done one
way repeated measures ANOVA's with a Dunnetts comparing to my baseline
measurement to describe whether there werre changes within a specific group,
and compared individual groups using a two way ANOVA (I am obviously not a
statistician). Reading a few books I get the impression I should be using a
form of (M)ANOVA combined with orthogonal polynomial contrasts (?multiplying
a matrix of my observations at time points by a matrix of contrasts and
using the resultant matrix as by dependent variable). Am I completely way
off line? I would be grateful for any help or directin towards a good text
book. I have access to S-plus but unfortunately not SAS.
Thanks
Simon
|