Comparison Between Three Groups Over Time

#1
Which statistical test should I use to compare between three groups over time (28 days) using means calculated for the whole group?
 

obh

Active Member
#4
Hi Ketone,

What are the three groups?
Do you compare the three groups? or each group to the same group in different situation? (like before taking the medicine, 1 day after medicine. 5 days after medicine)
 
#5
Hi Ketone,

What are the three groups?
Do you compare the three groups? or each group to the same group in different situation? (like before taking the medicine, 1 day after medicine. 5 days after medicine)
Hi Obh, I have one control and two other groups receiving interventions. I have to record the average of daily food intake of each group and compare it with the average food intake of other groups over a period of 28 days to find out if and when the difference becomes significant.
 

obh

Active Member
#6
So you have the daily food for each of the 28 days?
How many interactions do you have?
Like interactions in days: 3,11,14,25?
 

obh

Active Member
#8
What are the differences between the 3 groups? I assume the control didn't have any interaction? What is the difference between the two groups?
 
#10
What are the differences between the 3 groups? I assume the control didn't have any interaction? What is the difference between the two groups?
The food intake for one group started to decrease after one week for one intervention group and after two weeks for the other group as compared to the control group.
 

obh

Active Member
#11
So there is only one change? or 4?
If 4 is it always 1 week after?

Example:

Control: 50, 50, 50, 50, 50, 50, 50, 50,
Group1: 50, 40, 40, 40, 30, 30, 30, 30
Group2: 50, 50, 40, 40, 40, 30, 30, 30

If so, do you expect to get the same result on Group1 week2 (n) and group2 week3 (n+1)?

What do you compare between the groups?
 
#12
So there is only one change? or 4?
If 4 is it always 1 week after?

Example:

Control: 50, 50, 50, 50, 50, 50, 50, 50,
Group1: 50, 40, 40, 40, 30, 30, 30, 30
Group2: 50, 50, 40, 40, 40, 30, 30, 30

If so, do you expect to get the same result on Group1 week2 (n) and group2 week3 (n+1)?

What do you compare between the groups?
I think I did not understand the point about interaction and wrote 4 as in number of weeks involved. The change is decrease in food intake(g/rat/day) and that is what I want to compare, i.e, which intervention decreased the food intake more and was the decrease significant. According to my results, the food intake is almost the same in three groups for first week then it starts decreasing for group 1 in the 2nd week, while that of control and group 2 is almost the same and in the 3rd week, the food intake for group 3 also starts to decrease as compared to the control. What I want to know is that is it possible to statistically compare food intake "between" the groups each day? I am sorry for not making much sense but I am fairly new to statistics and I am trying to learn. Thank you so much for helping me out.
 

obh

Active Member
#13
The one way ANOVA, may compare if the difference of the averages of sample groups is significant.

But I still don't understand the research.
What are the differences between the 3 groups? say what cause to the change in the intake?
 
#14
The one way ANOVA, may compare if the difference of the averages of sample groups is significant.

But I still don't understand the research.
What are the differences between the 3 groups? say what cause to the change in the intake?
The interventions are olfactory stimulation with two different odorants for group 1 and 2 and these odorants reduce appetite in rats. One of the measures of appetite that I have chosen is food intake. I want to fcompare whether the reduction in appetite (food intake) is significant and at which point during the course of intervention does the difference in food intake becomes significant. I don't know how to incorporate in and to account for the time factor with one way ANOVA.
 

obh

Active Member
#15
On one way ANOVA you run a test separately for each day.
I assume the reduction increase over time? and then become constant? then you can run the test only on the last day to see id the difference is significant. (if the test power is strong enough, otherwise you may need to group days).

If it is significant for the last day and insignificant for the first day you may need to continue day2 day3 until getting a significant result.
But since you run multiple tests you may get significant results in day 5 and insignificant results in day6.
 
#16
On one way ANOVA you run a test separately for each day.
I assume the reduction increase over time? and then become constant? then you can run the test only on the last day to see id the difference is significant. (if the test power is strong enough, otherwise you may need to group days).

If it is significant for the last day and insignificant for the first day you may need to continue day2 day3 until getting a significant result.
But since you run multiple tests you may get significant results in day 5 and insignificant results in day6.
Ok, thank you, I'll try to do that.