Statsitics, Research Methods, SAS, R, HLM, Rasch model
HLM (multilevel models) and econometric analyses (e.g., time series analysis, ARIMA, etc.) are treated as different approaches (the goal of which is to deal with data dependency problem), they can be implemented in the same model via. SAS PROC GLIMMIX. However, I believe doing this is computationally demanding and models may not converge.
For detailed discussion of this topic, see p.28 of my HLM manual:
I will edit this essay later.
Imagine I have an HLM model where level-1 are students and level-2 are schools. I can enter teacher-level variables, but people who are used to HLM software by SSI will wonder how the use of teacher-level variables is possible without making the model 3-level models (level1 students, level2 teachers, level3 schools). This is because in the old HLM tradition (people who studied HLM using HLM software in 1990s), equations are written/processed separately by levels and data must be prepared by levels (student data file and school data file).
If there are student-level and school-level, some HLM users may wonder how teacher-level variables can enter the model (or prepared in what file? student-level file or school-level file???).
People who use SAS or other software programs (maybe R?) will not wonder this because they think in terms of one whole equation and say, "of course we can enter variables of any levels."
The software programs they use adjust degree of freedom to take into consideration that teacher-level variables have a lot less number of values compared to the outcome variable.
K-R option in PROC GLIMMIX adjusts degree of freedom to account for the fact that group-level variables have a lot less possible values when compared to outcome variables.
K-R degree of freedom option seems most appropriate for multilevel modeling applied in educational evaluation studies (where typically students are nested within schools). This option kicks in only when at least one random coefficient is estimated (e.g., intercepts as random effects).
proc glimmix data=i3d;
class school ;
model y =< here variable names >
/solution ddfm=kr dist=normal link=identity s ;
random int /sub=school;
run;
After playing with data type of simulation (I will do a better and well-documented simulation in the future), I learned the following:
For student-level (level-1) variables, degree of freedom under KR option is close to the number of students (minus a small number of cases maybe close to the number of predictors). DF is larger if variance contained in the variable is greater. For binary variables, DF is the largest when variance is .50 (DF gets smaller as a proportion gets close to 0 or 1).
For school-level (level2) variables, DF is close to the number of schools minus a small number maybe close to a number of predictors. This may be also adjusted by variance of the variable.
I created a fake teacher-level predictors by creating two possible numeric values per school. DF for this variable was close to the number of teachers in the data (two per school) minus a small number. I think this is also adjusted by variance of the variable (which I didn't test exactly).
PROC MIXED and PROC GLIMMIX produce identical results for the following linear model settings.
proc glimmix data=kaz.asdf;
where IMPACT_ANALYSIS=1;
class school;
model
y =
x1 x2 x3
/solution ddfm=kr dist=normal link=identity s ;
random school;
run;
proc mixed data=kaz.asdf;
where IMPACT_ANALYSIS=1;
class school;
model
y =
x1 x2 x3
/solution s ddfm=kr;
random school;
run;
When you run a statistical interaction model (e.g., Y=TREATMENT + GENDER + TREATMENT*GENDER), you also want to run a mathematically equivalent model:
Y= T_MALE + C_MALE + T_FEMALE + C_FEMALE.
SUBGROUP defines the four groups below.
This is for linear model (HLM because I have the random statement).
proc glimmix data=asdf3 namelen=32;
class CAMPUSID SUBGROUP;
model y=
x1 x2 x3
SUBGROUP
/solution ddfm=kr dist=normal link=identity s noint;
lsmeans SUBGROUP / ilink diff;
random int / subject = CAMPUSID;
ods output ModelInfo=x1var1 ParameterEstimates=x2var1 CovParms=x3var1
nobs=x4var1 Diffs=DIF_RESULT1;
run;
This is for logistic regression model.
proc glimmix data=asdf METHOD=RSPL;
class CAMPUS_14 subgroup;
model y=x1 x2 x3 subgroup
/dist=binomial link=logit s ddfm=kr;
lsmeans group / ilink diff;
ods output ModelInfo=x1var1 ParameterEstimates=x2var1 CovParms=x3var1
Diffs=DIF_RESULT1 LSMeans=LS1;
run;
When running PROC GLIMMIX (SAS) in a macro-driven way (e.g., running similar models 100 times), what gets annoying is some HLM models do not converge and you have to comb through output and decide which models to convert to fixed effect models, which is simpler and is easier to converge. The following allows the execution of a fixed model (non-HLM) when a random effect model (HLM) fails.
The following macro (%mend checkds;) checks if the first random effect model produces one of the result files (parameter estimates) and if it doesn't exist (i.e., random effect model did not converge), it will run the model without the random effect statement.
proc glimmix data=asdf METHOD=RSPL;
class CAMPUS_14;
model &out = &main
stud_char
interX
&predictors
/dist=binomial link=logit s ddfm=kr STDCOEF;
random int / subject = CAMPUS_14;
covtest /wald;
ods output
ParameterEstimates=kaz1
CovParms=uekawa1
ModelInfo=estes
dimensions=diminfo
ConvergenceStatus=concon
FitStatistics=FITSTAT
;
run;
data hlm1;
hlm1="HLM ";
run;
/*Check if converged and if not run fixed model*/
%macro checkds(dsn);
%if %sysfunc(exist(&dsn)) %then %do;
/*there is concon created*/
%end;
%else %do;
/*delete imcomplete data from the previous proc
that did not converge*/
proc datasets;
delete kaz1 estes diminfo concon FITSTAT hlm1;
run;
proc glimmix data=asdf METHOD=RSPL;
class CAMPUS_14;
model &out = &main
stud_char
interX
&predictors
/dist=binomial link=logit s ddfm=kr;
ods output
ParameterEstimates=kaz1
/*CovParms=uekawa1*/
/*nobs=jeana */
ModelInfo=estes
dimensions=diminfo
/*ConvergenceStatus=concon*/
FitStatistics=FITSTAT;
run;
data hlm1;
hlm1="Fixed";
run;
%end;
%mend checkds;
/* Invoke the macro, pass a non-existent data set name to test */
*%checkds(work.concon);
*%checkds(work.uekawa1);
%checkds(work.FITSTAT);
covtest /wald;
proc glimmix data=qc2 METHOD=RSPL;
class group_ID;
model Y = X
/dist=binomial link=logit s ddfm=kr STDCOEF;
run;
It is not clear how coefficients are standardized. Based on my investigation, centering is definitely done around the variable's grand mean; however, SD used for standardization is not 1. When I simulated it, SD was around 0.036 (meaning I created a z-score using mean=0 and STD=0.036 to get the same coefficient off STDCOEF,
proc glimmix data=temp2 /*Method=RSPL*/ ;
class CAMPUS;
model Y=/dist=binomial link=logit s ddfm=kr;
random int / subject = CAMPUS_14;
covtest /wald;
output out=gmxout_alglog residual=resid;
RUN;
Tips and Strategies for Mixed Modeling with SAS/STAT® Procedures Kathleen Kiernan, Jill Tao, and Phil Gibbs, SAS Institute Inc., Cary, NC, USA
ABSTRACT Inherently, mixed modeling with SAS/STAT® procedures, such as GLIMMIX, MIXED, and NLMIXED is computationally intensive. Therefore, considerable memory and CPU time can be required. The default algorithms in these procedures might fail to converge for some data sets and models. This paper provides recommendations for circumventing memory problems and reducing execution times for your mixed modeling analyses. This paper also shows how the new HPMIXED procedure can be beneficial for certain situations, as with large sparse mixed models. Lastly, the discussion focuses on the best way to interpret and address common notes, warnings, and error messages that can occur with the estimation of mixed models in SAS software.
http://support.sas.com/resources/papers/proceedings12/332-2012.pdf
glimmix data=xxx METHOD=RSPL ITDETAILS;
class xxx;
model xxx= xxx
/dist=binomial link=logit s ddfm=kr;
random int / subject = xxx;
NLOPTIONS MAXITER=100;
run;
Output Added:
-------------
Name: ModelInfo
Label: Model Information
Template: Stat.Glimmix.ModelInfo
Path: Glimmix.ModelInfo
-------------
Output Added:
-------------
Name: ClassLevels
Label: Class Level Information
Template: Stat.Glimmix.ClassLevels
Path: Glimmix.ClassLevels
-------------
Output Added:
-------------
Name: NObs
Label: Number of Observations
Template: Stat.Glimmix.NObs
Path: Glimmix.NObs
-------------
Output Added:
-------------
Name: Dimensions
Label: Dimensions
Template: Stat.Glimmix.Dimensions
Path: Glimmix.Dimensions
-------------
Output Added:
-------------
Name: OptInfo
Label: Optimization Information
Template: Stat.Glimmix.OptInfo
Path: Glimmix.OptInfo
-------------
Output Added:
-------------
Name: IterHistory
Label: Iteration History
Template: Stat.Glimmix.IterHistory
Path: Glimmix.IterHistory
-------------
NOTE: Convergence criterion (GCONV=1E-8) satisfied.
NOTE: At least one element of the gradient is greater than 1e-3.
Output Added:
-------------
Name: ConvergenceStatus
Label: Convergence Status
Template: Stat.Glimmix.ConvergenceStatus
Path: Glimmix.ConvergenceStatus
-------------
Output Added:
-------------
Name: FitStatistics
Label: Fit Statistics
Template: Stat.Glimmix.FitStatistics
Path: Glimmix.FitStatistics
-------------
Output Added:
-------------
Name: CovParms
Label: Covariance Parameter Estimates
Template: Stat.Glimmix.CovParms
Path: Glimmix.CovParms
-------------
Output Added:
-------------
Name: ParameterEstimates
Label: Solutions for Fixed Effects
Template: Stat.Glimmix.ParameterEstimates
Path: Glimmix.ParameterEstimates
-------------
Output Added:
-------------
Name: Tests3
Label: Type III Tests of Fixed Effects
Template: Stat.Glimmix.Tests3
Path: Glimmix.Tests3
-------------