Repeated measures analysis using mixed model - is this set up to yield the results I hope for?
Dec 8, 2019 12:35 PM(461 views)
Help! I’m trying something new for me using JMP Pro 14 and I’m not sure if I have set up the analysis (Fit Model) correctly to answer my questions.
I have a repeated measures experiment where the number of eggs laid by 237 individual insects was counted in 8 consecutive 30-min time intervals. The different insects were evaluated over the course of 29 different replicate days
It was later found that 68/237 of the insects had ruptured ovaries; the other 169 had intact ovaries. This may affect the pattern of egg release; thus the insects were classified by their rupture status before analysis of egg laying.
I would like to know whether the ruptured state affected egg laying.
I would like to compare eggs per interval separately for the ruptured and the unruptured groups using a procedure like LS Means Tukey HSD; however, I cannot get the sets of eggs per time interval means for the ruptured and unruptured groups individually.
I would like to know if there was an interaction between rupture status and egg laying across intervals.
I would also like know if my replicate days were significant (I don’t expect them to by different); I’m not sure if I have accounted for it correctly in this model.
I have my data in a stacked format. Here is the Model Specification
Thanks in advance for advice and correction...I know JMP can do what I need, but is what I have designed appropriate for my needs?!
I have not checked in detail, but just a comment on the personality in Fit Model. As you have JMP Pro, you might want too check out Mixed Models personality instead of the LS approach with REML. THere you would also have options to take care of different covariance structures. May someone else can chime in to make more comments on your specific example.