Pre and post- test on more then one day

#1
Hey.
i have a data setup, where i don't really know which test to use.

so i have 6 animal with an electrode in the brain. on day zero we measure the baseline of pain.
on day one we start stimulating the animals.

so we make a pre- and posttest on day 1, on day 3 and day 5 (pre and post stimulation)

so i want to know if there is a general difference between pre- and post tests, but not if there is for example a bigger difference between pre and post tests on day 1 than on day 3


so the design looks like this:
Unbenannt.PNG


Thank you

Marco
 

j58

Active Member
#2
You can used a linear mixed model with animal as a random variable, and each pre-post difference as the dependent variable. I don't know what the difference is between "baseline" and "pre_1."
 

hlsmith

Omega Contributor
#3
Yes, please better define these data in above table. Do you really need the Baseline and are you turning the electrode on and off?
 
#4
You can used a linear mixed model with animal as a random variable, and each pre-post difference as the dependent variable. I don't know what the difference is between "baseline" and "pre_1."
Yes, please better define these data in above table. Do you really need the Baseline and are you turning the electrode on and off?
sorry forgot it.

so we measured the baseline of pain-threshold. after that we damaged a nerve of the rats. then we measured again the pain threshold before and after stimulating with the electrode.
so the pain threshold pre-stimulation is much lower than the baseline, and is supposed to increase after the stimulation.

So the baseline is just measured to see if the surgery actually decreased the pain threshold, and the important question is: does the stimulation (from pre- to post test) increase the pain threshold.

thank you
 

hlsmith

Omega Contributor
#5
It seems unconscionable and inhumane to hurt a living creature and not even have a solidified analytic plan. What if the experiment isn't well powered, so you cannot even generate conclusive results either way. Do you not report to an ethics or institutional review board with an a priori study protocol?
 
#6
It seems unconscionable and inhumane to hurt a living creature and not even have a solidified analytic plan. What if the experiment isn't well powered, so you cannot even generate conclusive results either way. Do you not report to an ethics or institutional review board with an a priori study protocol?
this is a theoretical protocol we use in a statistic-cours in my university ....