Sas proc corr

saza — Thu, 17 Feb 2022 21:49:00 GMT

So I think I learned more about stats from this forum than my actual class and now I just want to apply what I learned through proc corr to different variables. Yesterday I was using proc reg when I later realized I should have used proc corr for this question:

Is there a relationship between gestation and baby weight? Yes, positive relationship

State your null and alternative hypothesis:

Null hypothesis:

(Where

is the population slope)
Alternative hypothesis:

Level of significance:
What is the rho value for this analysis? Rho=0
Make your scatter plot for this analysis, does it confirm your rho findings? Yes, I believe it does but not sure if the scatterplot proves it.
Does a statically significant relationship exist between these two variables? If so, interpret:
Yes, since the p-value of the model in the ANOVA table is p=0.001 <0.05 significant, we conclude there is a linear relationship.

I guess what I want to know if i'm making sense with this code and the questions asked or is there another way to do this?

I used this code:


proc corr data=birthw.birthweight;
var gestation birthweight;
run;

Re: Sas proc corr

Reeza — Thu, 17 Feb 2022 22:05:02 GMT

If you're testing slope, regression was the correct method.
If you're testing correlation, proc corr is correct and your hypothesis doesn't align.

Both of these are methods that are taught, the actual correct approach depends on data and data usage.
But the answer as to which is the right answer your professor is expecting - so which ones have you been taught as that's likely what your professor is expecting.

topic Re: Sas proc corr in Statistical Procedures

Sas proc corr

Re: Sas proc corr