If we haveΔR2 in our table, does the R2 value for Step 1 go in the table or the note? Could the R2 value technically be seen as a change from the null? Or is that an improper understanding of it?
Definately wait for the TA's reponse, but my understanding is that you would not put the R^2 for step one in the table, but you would list it in the note. Or you could put the R^2 and both delta R^2s all in your note...
I also think that a change from the null is stated with the significance and it wouldn't really work to state a change in R^2 for step one as this would imply it changed from some other step...not from the null model...but, again, wait for the TAs!
I am asking Sung about how you would report the R statistics in a table Because this seems redundant to me. But I will let you know when he answers.
However, i CAN answer your second question.
You are correct in stating that the first step in the model is testing its difference from the null model. You do not report change statistics for your first model, because it is the model you are controlling and which will then be changed in the next steps/blocks.
Wanna test this? Check out your R^2 stat versus your R^2 change stat for your first model...then you can also check out the F-stats in the anova table and the change statistics table for your first model ...compare...What do you see?
3 comments:
Definately wait for the TA's reponse, but my understanding is that you would not put the R^2 for step one in the table, but you would list it in the note. Or you could put the R^2 and both delta R^2s all in your note...
I also think that a change from the null is stated with the significance and it wouldn't really work to state a change in R^2 for step one as this would imply it changed from some other step...not from the null model...but, again, wait for the TAs!
I am asking Sung about how you would report the R statistics in a table Because this seems redundant to me. But I will let you know when he answers.
However, i CAN answer your second question.
You are correct in stating that the first step in the model is testing its difference from the null model. You do not report change statistics for your first model, because it is the model you are controlling and which will then be changed in the next steps/blocks.
Wanna test this? Check out your R^2 stat versus your R^2 change stat for your first model...then you can also check out the F-stats in the anova table and the change statistics table for your first model ...compare...What do you see?
Hope it's as cool to you as it is to me!
Post a Comment