Repetead Intervention : which is the right test?

#1
Hi everybody,
I'm new with this forum, hence sorry for any mistakes.

I'm running a protocol with this scenario:

ROUND 1: Baseline Data (T0) - Exercise - Data T1 (immediately after) - Data T2 (after 24h) - Data T3 (after 48 h)
ROUND 2: After 1 week Data (T5) - Exercise - Data T6 (immediately after) - Data T7 (after 24h) - Data T6 (after 48 h)


I would like to know which test can I use to compare the 2 rounds. I'm interested in understanding if the repetition of the Exercise (at 1 week of distance from the first one) change somehow the variable I'm measuring in the timepoints after the exercise.

Let me know if I was clear or maybe you need more information.

Thank you in advance,
Rosaria:wave:
 

Karabiner

TS Contributor
#2
Please explain the variable you are measuring
in more detail. And please report your sample size.

Moreover, could you explain what you precisely
mean by
understanding if the repetition of the Exercise (at 1 week of distance from the first one) change somehow the variable I'm measuring in the timepoints after the exercise.
Do you want to know whether single time points
differ(t2 from t1, t6 from t5 etc.)? Or whether
the according time points within the two rounds
differ (ie. t1 ves. t5, t2 vs. t6 etc.)? Or whether the
differences differ (e.g. is t1-t2 different from t5-t6)?
r whether any of the time points (t5, t6, t7, t8) in the
second round differs from t4? Or something else?

With kind regards

K.
 
#3
Ok, I'll try to explain it better.
I have 12 subjects and I have measured continuos variables (ie: kN of maximum voluntary contraction of muscles)

I want to know the impact of the repetition of the exercise on the variables analyzed, that means to evaluate if the changes of the variables in the first week (ie: the maximum voluntary contraction of muscles is reduced immediately after the exercise, and than goes back to the baseline values after 48 hrs) are significantly different when compared with the changes in the second week (ie: the maximum voluntary contraction of muscles is reduces immediately after the exercise, but less than the first week).

Does it make more sense now?

Thank you!
 

Karabiner

TS Contributor
#4
If I understand correctely, then a general test
would need 2 factors, round (2 levels) and
time (4 levels). You could use a
repeated measures ANOVA with these two
factors as within-subject factors. The
interaction between round and time
would then tell you whether the
differences between baseline/immediate/24h/48h
differ between rounds.


the maximum voluntary contraction of muscles is reduces immediately after the exercise, but less than the first week
If you are only interested in this specific question,
then you could use just baseline & immediate as levels
for the time factor.

With kind regards

K.
 
#5
Great! I think I'm on the right way.
I did what you suggested me and now the interaction round*time resulted statistically significant. But this is the overall effects, isn't it?

Is there a way to assess per each couple of time points if they are statistically significant different? Something like the couple comparison that I see among time?

Thank you and regards.
 

Karabiner

TS Contributor
#6
You could consider comparisons using dependent
samples t-tests, ususally with some kind of correction
for multiple testing. In case of Bonferroni correction
the new alpha would be alpha/4 (4 comparisons).

With kind regards

K.