SPSS interpreting change scores, am I doing it right??

#1
I had to calculate some change scores for my dissertation. I didn't even know what they were until my prof told me to do that in addition to my other statistics. I tested 2 groups, 1 got supplements the others a placebo and tested them before and after taking them. They filled in mood and stress questionnaires and I got only very slightly significant results for stress everything else is not significant but goes into a certain direction. So my professor told me to calculate the change scores to look more indepth at what is happening.

I was unable to find anything about how to interpret them online or in my stats book and my prof is not replying my email. I for example analyzed the change score for stress between the two groups and got significant results. But what does that mean now? That the supplements probably had a positive effect? Or that there is a difference between placebo and supplements? or should I look at the change score for the supplement group individually and for the placebo group and see if each one of those is significant or not? But if it is, what does it actually tell me?
 

hlsmith

Omega Contributor
#2
Can you define what you mean by change score. Are you referencing the difference between pre and post scores? Randomization really helps, but you should also ensure baseline values are comparable across the two groups. Also, could loss-to-follow-up happen in your study?
 
#3
A difference score is a variable that has been formed by subtracting one variable from another.
i.e., DIFFSCORE = VAR1 - VAR2.
A difference score is valid to the extent to which it actually measures what you intend to measure.