Cherry. Im kinda happy I finally understand it conceptually. It took me a while. It's just helping repeating back what I learn so I can remember...
Soo in short.
Null hypotheses are that its 1) normaly distributed 2) there is no difference/correlation.
when alpha is .05 we can do this:
If p>.05 then are null hypthesis is right. We can accept the data as is theres no problems with it. we accept the null hypothesis
If p<.05 than than it's not right. We can't say isn't normally distributed, nor can we say that theres no correlation. we rject the nlul hypothesis.
After rejecting it we can do a regression to see what the correlation is. If the p value is small for a correlation cooeficient it belongs in our equation. If it's high, then there was no correaltion like we assumed and we just set the correlation coefficient = 0.
Ok Im done. ughhh kinda nervous.