F Test for Least Squares Means Estimates formula

[Divinesking]

Hello everyone,

I am trying to understand how the joint F value from proc mixed using lsmestimate option is being calculated. An example data, sas

data test;
input Y visitno subid;
datalines;
7 1 1
9 1 2
8 1 3
7 2 1
8 2 2
6 2 3
;

proc mixed data=test ;
class visitno subid;
model y = visitno/solution outp=pred ;
repeated visitno /type=un subject=subid;
*contrast 'visitno' [1 1], [1 2];
lsmestimate visitno [1 1], [1 2]/joint(chisq) cl e;
run;

Least Squares Means Estimates 
Effect Label Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper 
visitno Row 1 8.0000 0.5774 2 13.86 0.0052 0.05 5.5159 10.4841 
visitno Row 2 7.0000 0.5774 2 12.12 0.0067 0.05 4.5159 9.4841 



F Test for Least Squares Means Estimates 
Effect Num DF Den DF Chi-Square F Value Pr > ChiSq Pr > F 
visitno 2 2 228.00 114.00 <.0001 0.0087 

Can someone help me understand how the F Value of 114 is derived?

Thank you for your help.

Best,

Divine

[PaigeMiller]

Please click on the {i} icon and paste the output into the window that appears. This will maintain the formatting of your output, and make it readable.

[PaigeMiller]

The Chi-squared value is computed to be 228, so the F-value is 228/2 = 114.

For future reference, I wanted you to copy the TEXT from the LISTING window and paste it into the {i} window. If you copy the HTML, it still isn't very readable. If you don't have a LISTING window, then a screen capture of the HTML is fine.

[Divinesking]

Thank you PaigeMiller. 

So to ask but how to calculate the chi-square value of 228 with the joint(chisq) option. Is it the ratio of the independent chi-square?

Best,

Divine

[PaigeMiller · Accepted solution]

https://documentation.sas.com/?cdcId=pgmsascdc&cdcVersion=9.4_3.4&docsetId=statug&docsetTarget=statu...

[Divinesking]

Thank you again!