Reply With QuoteDoing a multiple regression with the three change scores seems reasonable to me....why doesn't your committee want you to compute the change scores?
I am working on my dissertation proposal and have run into a snag in the analyses section. I am trying to analyze data drawn from two time-points. I would like to determine whether changes in one variable from time 1 to time 2 will be positively related to changes in a second variable from time 1 to time 2, even after controlling for changes in a third variable from time 1 to time 2. (All variables are continuous.)
My research advisor has suggested that I use a multiple regression, but I am unsure of how one would do this without first computing the three change scores (which my committee doesn't want me to do). My first thought was that I would need to conduct some type of time series analysis to see the relationship between the changing values.
I was hoping that someone could advise me regarding whether I can conduct the analyses as a regression using the original six values, or if I shoud be using another type of analysis instead (and if so, could point me in the right direction).
Thank you in advance for any assistance you can provide!
Doing a multiple regression with the three change scores seems reasonable to me....why doesn't your committee want you to compute the change scores?
I asked my advisor if I could do it that way, and she said that that's an old-fashioned way to conduct analyses, and that it's not a sound methodology. Another member of my committee agreed with her, as well.
Conceptually, it makes sense to me to do it that way, but it's not ultimately my decision.
Thanks for your help.
Posting Permissions
|
|