The Student Room Group
Uni students - Complete our survey >>
Uni students - Complete our survey >>
Uni students - Complete our survey >>

Null-hypothesis

Having a hard time figuring out this question: ''which null-hypothesis is interesting to test in simple linear regression''. Anyone brighter than me? Pleaase! :smile:
Reply 1
Avatar for medhelp
medhelp
14
I think simple linear regressions is basing predictions on x = what you expect to get from y


I think it's just trying to get you to come up with a null hypothesis based off the above model. I'm guessing this from AS psychology?
Reply 2
Avatar for Muminmamman
Muminmamman
OP
Original post by medhelp
I think simple linear regressions is basing predictions on x = what you expect to get from y


I think it's just trying to get you to come up with a null hypothesis based off the above model. I'm guessing this from AS psychology?


It's from a statistics course at University. I'm finding this very difficult, not sure I'll figure it out but thanks for your help!
Reply 3
Avatar for medhelp
medhelp
14
Original post by Muminmamman
It's from a statistics course at University. I'm finding this very difficult, not sure I'll figure it out but thanks for your help!


I'm crap at stats lol
maybe there's certain types of hypothesis testing for certain regression lines idk anyway good luck finding out what is it :smile:
A null hypothesis is when you predict that there is no difference or relationship between two factors or co variables
A linear regression graph is a straight line graph
So by deduction I'd assume that a null hypothesis linear graph would be the one with the most scattered points around the best for or low correlation coefficient
Hope that helps!
Doing psychology at uni and did a stats module at alevel
*fit

Quick Reply

Related discussions

Latest

Trending

Trending

Articles for you