Hello again chicos and chicas,

So we went through almost all the important stuff for multiple linear regression, but I have one more for you and then I think we might go further into developing beautiful word of analysis and statistics 😊. So last for now, but not least of course:

Model fit verification

So as in simple linear regression we compute RSE and R² for our model – here link to my post explaining that.

R² close to 1 will again indicate that the model explains a lot. But what we need to remember, that R_­­² will always increase when more variables will be added to the model, even if they are not so significant. In order to see that better, we shall consider examples, so we will, but in one of the following posts.

Okay, what about RSE? So formula for that is in general given by: (RSS/(n-p-1))^1/2, which in simple linear regression case simplifies to the formula given previously. And again, deciding what are the outcomes of RSE requires knowledge about the model itself and its values.

That was quick. There are many things about the model that needs to be considered except that, but for the introduction theory, I would say that is it and further important details shall be discovered on ‘living’ examples.

Thank you and see you soon,

xoxo,

szarki9