|
|
Comparison of two R-squared |
|
Message/Author |
|
|
I need to compare two path analysis models with different predictors with regard to the amount of explained variance in one categorical outcome variable Y. The first model (restricted model) contains one continuous predictor X1, the second (unrestricted) model includes a second predictor X2, which functions as a mediator (X1 -> X2 -> Y). Compared to the restricted model, R-squared in the unrestricted model is higher (by 0.07). Is there any way to test this increment for significance? |
|
|
I don't know of any such test. I would check a regression text to see if this is discussed there. |
|
Back to top |
|
|