topic Re: F Test in SAS Studio

F Test

Lilo — Wed, 27 Jan 2021 18:15:08 GMT

Hello,

I am trying to find the appropriate code to run a F test to test a hypothesis but so far no luck

Here's the data:

data reghwk4; 
input X Y;
datalines;
29.7 16.6 
68.4 49.1
120.7 121.7
217.2 219.6
313.5 375.5
419.1 570.8
535.9 648.2
641.5 755.6
;
run;

This is what I have tried so far:

proc ttest data= reghwk4 / h0=0 sides=1;
class X;
var Y;
run;

proc anova data=reghwk4;
class Y X;
model Y=X;
run;

proc reg data= reghwk4;
model y=x;
run;

Re: F Test

PeterClemmensen — Wed, 27 Jan 2021 18:24:39 GMT

What is your hypothesis?

Re: F Test

Lilo — Wed, 27 Jan 2021 18:28:03 GMT

Hypothesis: Ho equals 0 against H1 does not equal 0

Re: F Test

ballardw — Wed, 27 Jan 2021 18:32:01 GMT

Does your actual data have any repeats of values?

You probably want ANOVA but without Y as a class variable. But when there are no repeat values for the Class variable (or combinations) then the dependent variable has no variance. So the test doesn't report any results for the tests.

See this:

data reghwk4; 
input X Y;
datalines;
29.7 16.6 
68.4 49.1
120.7 121.7
217.2 219.6
313.5 375.5
419.1 570.8
535.9 648.2
641.5 755.6
29.7 16. 
68.4 49.
120.7 121.
217.2 219.
313.5 375.
419.1 570.
535.9 648.
641.5 755.
;
run;

proc anova data=reghwk4;
   class  X;
   model Y=X;
run;
quit;

Now there are multiple values for Y for each of the X's an so some variance to analyze.

You want to end Proc Anova with a Quit statement. It is one of the procedures you an use interactively with multiple model and optional statements and will be considered as "running" until a quit is supplied or another procedure runs.

Re: F Test

PaigeMiller — Wed, 27 Jan 2021 18:34:00 GMT

Are you comparing mean of X to the mean of Y? It would really be good if you state that clearly. Are you doing a "paired t-test" or an "unpaired t-test" (which is called "Independent Sample t-test" at that link). None of these produce an F-test.

Are you testing the linear relationship between X and Y? It would be good if you state that clearly. It's not clear why you are using a CLASS statement when none of your variables are CLASS. Please explain.

Re: F Test

PaigeMiller — Wed, 27 Jan 2021 18:36:55 GMT

Hypothesis: Ho equals 0 against H1 does not equal 0

This explains nothing. What statistic are you comparing to zero? Means? Standard Deviations? Kurtosis?

Statistic is equal to zero on what variable? X or Y or both or something else?

Re: F Test

Ksharp — Thu, 28 Jan 2021 12:56:17 GMT

F test is for ANOVA . But you don't have CLASS variable .