topic Confidence Interval Intercepts with Lines in SAS Programming

Confidence Interval Intercepts with Lines

Ariel_1 — Tue, 02 Feb 2021 18:45:03 GMT

I was wondering if there was an easy way to determine where the confidence intervals of a line intercept two lines.

Example Below, I’d like to know where the confidence intervals of each line intercept 90 and 110. I'm on SAS 9.04

data data1;

input lot $ age results ;

datalines;

a 0 100

a 1 101

a 2 102

a 3 101

a 4 99

a 5 101

a 6 100

b 0 102

b 1 99

b 2 104

b 3 95

b 4 98

b 5 97

b 6 104

c 0 104

c 1 99

c 2 97

c 3 102

c 4 98

c 5 100

c 6 101

;

run;

proc glm data=data1;

class Lot;

model results = lot age*lot/ noint solution ;

output out= lsmeans lclm=lclm95 uclm=uclm95 p=p ;

run;

quit;

Re: Confidence Interval Intercepts with Lines

PaigeMiller — Tue, 02 Feb 2021 19:15:07 GMT

You could do it by seeing what values of AGE result in confidence limits of 90 and 110.

You could augment the data set with records where AGE varies slowly from 0 to 6 (let's say by 0.01 increments) and where the Y value (RESULTS) is missing so that the record is not used in the regression. Re-run the regression. Then examine the output data set to see where the confidence limits are very close to 90 and 110.

Re: Confidence Interval Intercepts with Lines

Ariel_1 — Wed, 03 Feb 2021 13:52:18 GMT

Paige - that's a very good idea. However, I was trying to program it since I have many variables that I am looking at.