HELP with experimental analysis (fear conditioning)

#1
I am running an olfactory fear conditioning procedure in which 2 odors (odor A/ odor B) are either paired with a footshock (CS+) or absence of footshock (CS-), and I have several drug treatment groups. So, each animal has a drug treatment assignment, a % time freezing score in response to CS+, and a % time freezing score in response to CS-. (see below for example data setup)

Treatment.......................freeze CS+ (%)..................freeze CS- (%)
dopamine antagonist...............85.....................................15
5HT antagonist.......................45.....................................55
GABA antagonist.....................65.....................................35


So I'm just wondering what the most appropriate statistical test would be to analyze these data. My thoughts are that it should be a mixed-model ANOVA as each animal has a between subjects variable (drug treatment) and 2 within subjects variables (freeze CS+, and freeze CS-). However, similar studies I've read have reported using a 2-way ANOVA. At the very least wouldn't it need to be a repeated measures ANOVA? (btw I'm using SPSS)
Thanks in advance for any help!!!:cool:
 

Karabiner

TS Contributor
#2
My thoughts are that it should be a mixed-model ANOVA as each animal has a between subjects variable (drug treatment) and 2 within subjects variables (freeze CS+, and freeze CS-).
Seems correct.
However, similar studies I've read have reported using a 2-way ANOVA.
Maybe these studies did not have repeated measures
(their subjects did not undergo several experimental
conditions).

With kind regards

K.