So I'm ready to run the stats of my thesis, and I'm a bit overwhelmed. My advisor only uses R, and I feel a lot more comfortable in SPSS (I have a couple guides, plus it's a million times more user friendly compared to R). My main issue is...I suck at stats. Math has always been my weakest subject. Also, I was all set to do multiple Ancovas for my analysis, but at the last minute, my committee changed it. Basically, I could use some guidance. Their suggestion was to do multiple linear regression models. I thought multiple regression at first, but they said no, multiple linear regression models.
To bring you guys up to speed, I'm looking at 4 different language outcomes in kids with autism before and after 2 different therapies were administered.
I'm going to have a single score for each language outcome. (score post therapy/score pre therapy) - 1). Obviously, these are my dependent variables.
Since the kids didn't all come from the same classroom, I'm using which classroom each child is in as a random effect.
I also have the age in months for each child, along with a teacher fidelity rating (to see how well the kids were actually getting one of the intervention methods).
And of course which therapy method each child is in as an IV.
Problem is, I don't even know how to get off the ground with this. Ancovas seems so much simpler...