topic Re: Problems when Absorbing two variables using Proc glm in Statistical Procedures

Problems when Absorbing two variables using Proc glm

Xudeer — Fri, 11 Aug 2017 01:23:32 GMT

Hi all,

I am running fixed effects model that regress outcome on A while controlling for both student FE (studentid) and course FE (courseid). I used Proc glm. My understanding is that i can absorb both studentid and courseid.
In my first model, I only absorbed one variable:
Proc glm data=data;
Class courseid;
Absorb studentid;
Model outcome=A courseid;
Run;

I had no problems running the model and the coefficients on A look correct. However when I add both course ID and student ID in the absorb function, sas fails to provide a valid coefficients for A. Shouldn't the two models-year-old exactly the same results?
Any insights are appreciated!

Re: Problems when Absorbing two variables using Proc glm

Xudeer — Fri, 11 Aug 2017 01:46:55 GMT

Re: Problems when Absorbing two variables using Proc glm

JacobSimonsen — Fri, 11 Aug 2017 08:17:48 GMT

Agree, it should give same estimates for A. Unless courseID is nested in A, because you then in practice also have absorbed A when you absorb courseID.

Re: Problems when Absorbing two variables using Proc glm

Xudeer — Mon, 14 Aug 2017 16:59:06 GMT

Thank you Jacob!

That is why I got confused. My courseID is not nested in A. A is actually teacher ID. The majority of courses are taught by multiple college instructors and each instructor is teaching multiple courses as well. The most confusing thing is that when I only absorb student ID and add courseID and my key variabe A (instructorID) as dummy variables, my model is totally fine:

SAS Output

Source DF Sum of Squares Mean Square F Value Pr > F Model Error Corrected Total

82000	32715.80481	0.39897	2.39	<.0001
241302	40251.64007	0.16681
323302	72967.44488

R-Square Coeff Var Root MSE second_2yr Mean

0.448362

118.6951

0.408424

0.344095

However, once I abosorb both studentID and courseID. the model explodes, incating that something is wrong:

SAS Output

Source DF Sum of Squares Mean Square F Value Pr > F Model Error Corrected Total

323250	72954.59899	0.22569	0.91	0.7029
52	12.84589	0.24704
323302	72967.44488

R-Square Coeff Var Root MSE second_2yr Mean

0.999824

144.4448

0.497028

0.344095

Re: Problems when Absorbing two variables using Proc glm

PaigeMiller — Mon, 14 Aug 2017 17:28:11 GMT

You have only showed us the code for one of the models. Your description seems to indicate a fairly straightforward change to the code for the second model, but it would still be nice if you showed it to us.

My other concern is that you have 82000 df for the model in the first output, this doesn't seem to be a likely number, it seems way too large for any type of teacher/testing scenario I am aware of. Also, the total degrees of freedom, over 300,000, is also way too large for any type of teacher/testing scenario. Can you explain why these numbers are so large?

Re: Problems when Absorbing two variables using Proc glm

Xudeer — Tue, 15 Aug 2017 00:47:07 GMT

Thank you for the response, PaigeMiller!

Here is my first model that only absorbs studentID:

proc glm data=derived.fouryear;
absorb student_nid;
class instructor_nid coursenid_ft;
model second_2yr=instructor_nid coursenid_ft / solution;
run;

Here are the outputs:

SAS Output

Dependent Variable: second_2yr

Source DF Sum of Squares Mean Square F Value Pr > F Model Error Corrected Total

81979	31583.90047	0.38527	2.25	<.0001
241323	41383.54441	0.17149
323302	72967.44488

R-Square Coeff Var Root MSE second_2yr Mean

0.432849

120.3472

0.414109

0.344095

Source DF Type I SS Mean Square F Value Pr > F student_nid instructor_nid coursenid_ft

68606	17514.98225	0.25530	1.49	<.0001
7867	10393.42479	1.32114	7.70	<.0001
5506	3675.49343	0.66754	3.89	<.0001

In the second model, everything remains the same, except that now I absorb courseID rather than having it as dummies:

proc sort data=derived.fouryear; by student_nid coursenid_ft;run;
proc glm data=derived.fouryear;
absorb student_nid coursenid_ft;
class instructor_nid;
model second_2yr=instructor_nid/ solution;
run;

Here are the outputs from the second model:

SAS Output

Dependent Variable: second_2yr

Source DF Sum of Squares Mean Square F Value Pr > F Model Error Corrected Total

323229	72948.94614	0.22569	0.89	0.7788
73	18.49874	0.25341
323302	72967.44488

R-Square Coeff Var Root MSE second_2yr Mean

0.999746

146.2955

0.503396

0.344095

Source DF Type I SS Mean Square F Value Pr > F student_nid coursenid_(IN ABOVE) instructor_nid

68606	17514.98225	0.25530	1.01	0.5041
254558	55420.96262	0.21771	0.86	0.8402
65	13.00126	0.20002	0.79	0.8338

Re: Problems when Absorbing two variables using Proc glm

PaigeMiller — Tue, 15 Aug 2017 11:02:05 GMT

You haven't addressed why there are so many degrees of freedom, this seems like an incredibly large number.

However, from the ABSORB documentation

Several variables can be specified, in which case each one is assumed to be nested in the preceding variable in the ABSORB statement.

So your two models are not equivalent.

Also, from the documentation

When you use the ABSORB statement, the data set (or each BY group, if a BY statement appears) must be sorted by the variables in the ABSORB statement.

Re: Problems when Absorbing two variables using Proc glm

Xudeer — Tue, 15 Aug 2017 22:33:18 GMT

Thank you PaigeMiller!

You haven't addressed why there are so many degrees of freedom, this seems like an incredibly large number.

- We have more than 300,000 observations (the data is student by course level transcript records from multiple cohorts of students from an entire four-year public college system)

However, from the ABSORB documentation

Several variables can be specified, in which case each one is assumed to be nested in the preceding variable in the ABSORB statement.

So your two models are not equivalent.

- I see. What I would like to do is to absorb studentID and courseID which are not nested within each other. Is there any model that SAS would allow that?

Also, from the documentation

When you use the ABSORB statement, the data set (or each BY group, if a BY statement appears) must be sorted by the variables in the ABSORB statement.

-- Yes, I sorted the variable before running the command.

Re: Problems when Absorbing two variables using Proc glm

JacobSimonsen — Wed, 16 Aug 2017 08:46:33 GMT

I dont think there is any procedure doing what you want. But, theoretically it is possible to "absorb" non-nested variables. As you maybe know, when using the absorb method, data and collumnvectors of the design matrix is projected into the orthogonal space of the design vectors defined by the variable(s) in the absorbstatement. This is quite simple if you have only one class variable in the absorb statement. If there are more variables (non-nested), then this projection becomes more complicated (in terms of calculation time). I experimented with this some years ago, and I didnt see any time efficient way to do it. So that is maybe the reason that it is also not possible with proc glm.

Re: Problems when Absorbing two variables using Proc glm

Xudeer — Wed, 16 Aug 2017 17:18:55 GMT

Thank you Jacob! For your information, STATA can absorb multiple non-nested variables, but it runs extremely slow for a large dataset such as mine.

I guess for SAS, I will have to only absorb only one variable while adding the other as dummies?