Hi -- I'm working through your resource, and it's great. Thanks for putting it together.
In 5.2.4 you say that "Public schools at mean ses have a -0.14-point decrease on average in math achievement relative to private schools". However, I think that this value is -.04859, or -.05. This is from the fixed effects estimate for "public", right? (if not, might be worth clarifying where this figure comes from)
While I have you, I also wonder about your interpretation of the data. You say that the great drop in level 2 (school) variance when you account for public/private makes sense as public/private is a level 2 variable. But the drop in variance actually appeared when you added level 1 individual student SES in the previous model, and it didn't really drop any further once you added school type.
I'm still getting my head around this stuff, but this makes sense to me -- controlling for SES, private/public doesn't have a great impact on maths, but controlling for individual SES will have an impact on the estimate of school importance, as students with similar SESs tend to cluster within schools.
As I say, I'm new to mlms, so I could be wrong on both counts. Thought you might appreciate evidence of someone engaging with your work either way though. Thanks again:)
Hi -- I'm working through your resource, and it's great. Thanks for putting it together.
In 5.2.4 you say that "Public schools at mean ses have a -0.14-point decrease on average in math achievement relative to private schools". However, I think that this value is -.04859, or -.05. This is from the fixed effects estimate for "public", right? (if not, might be worth clarifying where this figure comes from)
While I have you, I also wonder about your interpretation of the data. You say that the great drop in level 2 (school) variance when you account for public/private makes sense as public/private is a level 2 variable. But the drop in variance actually appeared when you added level 1 individual student SES in the previous model, and it didn't really drop any further once you added school type.
I'm still getting my head around this stuff, but this makes sense to me -- controlling for SES, private/public doesn't have a great impact on maths, but controlling for individual SES will have an impact on the estimate of school importance, as students with similar SESs tend to cluster within schools.
As I say, I'm new to mlms, so I could be wrong on both counts. Thought you might appreciate evidence of someone engaging with your work either way though. Thanks again:)