Got some questions on simple regression

#1
Hello everybody! I'm learning how to use SAS program and I have some questions when I got this ANOVA (view pic)

- Are the two parameters estimated significantly different from 0? explain your answer.
- Examine residual ... Is it a right/OK model?
- From the observed variability between weeks in the UK_Sessions, what percentage can be explained by the Spain_Sessions?

The first and the second questions are the most important for me.
Thank you so much.


https://i.stack.imgur.com/w0yfO.png
https://i.stack.imgur.com/8snSY.png
 

Dason

Ambassador to the humans
#2
Hi! :welcome: We are glad that you posted here! This looks like a homework question though. Our homework help policy can be found here. We mainly just want to see what you have tried so far and that you have put some effort into the problem. I would also suggest checking out this thread for some guidelines on smart posting behavior that can help you get answers that are better much more quickly.
 

hlsmith

Omega Contributor
#3
Those are some nice residual, but as Dason's is stating, you have to first show some work or interpretation yourself. We aren't just going to answer questions for you, if it is homework that could be unethical.
 
#4
Hi everybody,

First thank you for your time, I really appreciate that.
The funny thing is that i wrote the questions in the way that they look like homework but is wasn't my purpose. My homework is to learn SAS Base, so this is my explanation:

I got the data from an Excel sheet, import it to SAS, clean in and made scatter plot matrix, some graphs and the simple reg model which is
UK_Sessions = 202,2 + 0,34 * Spain_Sessions

Well, form my own I have the question "Are the two parameters estimated significantly different from 0? and why?"

My professor said that we only need to know the SAS code and the reg model but it would be great to know about the residual (this is for the 10 out of 10 I think :cool:) so I thought a good way to talk about the residual is if we can know if this is a good model only by seeing the residuals.

If you think this is only for homework and I don't want to work in it you are free to think it but I think this is over the homework, it's like studying more than the professor want me to do.

I will give you some SAS code and the work I've got until now.

proc CORR data=isdi.t outp=c;
var week sessions Spain_Sessions UK_Sessions Pageviews New_Visitor_S;
run;


View attachment 6435

proc corr data=isdi.t nosimple
plots=matrix (nvar=all histogram);
id week; * id for the toptip;
var week sessions Spain_Sessions UK_Sessions Pageviews New_Visitor_S;
ods graphics / imagemap=on width = 1000 height = 800;
run;


View attachment 6436
 

ondansetron

TS Contributor
#5
Let's start here.

1) What are some ways you could test the hypothesis that the parameters are individually different from zero (it only looks like you have one independent variable, so I'm assuming the intercept and the slope coefficient are the things they want you to test)?
2) List out the standard linear regression assumptions (at least, the ones pertaining to the random error term). Tell us how you could investigate whether or not these are satisfied (what kinds of graphs--with the axes defined).
3) Define Root MSE, R-squared, Adjusted R-squared, and Coefficient of Variation (CV). Which of these definitions will help you answer question 3?

Let us know what you think, and we'll be able to give you some guidance (we're more than happy to help)!