topic Re: when using proc mixed to treat subject as random effects or fixed effects, why get same result? in SAS Procedures

when using proc mixed to treat subject as random effects or fixed effects, why get same result?

annaliseshen — Fri, 13 Jan 2017 22:18:35 GMT

PROC MIXED DATA=TRY;
CLASS TREATMENT PERIOD SEQUENCE SUBJECT;
MODEL CONC=TREATMENT PERIOD SEQUENCE/SOLUTION;
RANDOM SUBJECT(SEQUENCE);
LSMEANS TREATMENT/PDIFF=control("A","B") CL ALPHA=0.1 ;
RUN;

PROC MIXED DATA=TRY;
CLASS TREATMENT PERIOD SEQUENCE SUBJECT;
MODEL CONC=TREATMENT PERIOD SEQUENCE SUBJECT(SEQUENCE)/SOLUTION;
LSMEANS TREATMENT/PDIFF=control("A","B") CL ALPHA=0.1 ;
RUN;

The SAS System

If WE treat subject(sequence ) as fixed effect

The Mixed Procedure

Model Information
Data Set	WORK.TRY
Dependent Variable	CONC
Covariance Structure	Diagonal
Estimation Method	REML
Residual Variance Method	Profile
Fixed Effects SE Method	Model-Based
Degrees of Freedom Method	Residual

Class Level Information
Class	Levels	Values
TREATMENT	2	A B
PERIOD	2	1 2
SEQUENCE	2	1 2
SUBJECT	10	001 002 003 004 005 006 007 008 009 010

Dimensions
Covariance Parameters	1
Columns in X	17
Columns in Z	0
Subjects	1
Max Obs Per Subject	20

Number of Observations
Number of Observations Read	20
Number of Observations Used	20
Number of Observations Not Used	0

Covariance Parameter Estimates
Cov Parm	Estimate
Residual	0.4089

Fit Statistics
-2 Res Log Likelihood	25.7
AIC (smaller is better)	27.7
AICC (smaller is better)	28.4
BIC (smaller is better)	27.8

Solution for Fixed Effects
Effect	TREATMENT	SUBJECT	PERIOD	SEQUENCE	Estimate	Standard Error	DF	t Value	Pr > \|t\|
Intercept					3.1780	0.4953	8	6.42	0.0002
TREATMENT	A				1.0800	0.2860	8	3.78	0.0054
TREATMENT	B				0	.	.	.	.
PERIOD			1		-0.1160	0.2860	8	-0.41	0.6956
PERIOD			2		0	.	.	.	.
SEQUENCE				1	-0.7100	0.6395	8	-1.11	0.2991
SEQUENCE				2	0	.	.	.	.
SUBJECT(SEQUENCE)		001		1	-0.1750	0.6395	8	-0.27	0.7913
SUBJECT(SEQUENCE)		002		1	-0.4000	0.6395	8	-0.63	0.5490
SUBJECT(SEQUENCE)		003		1	0.3850	0.6395	8	0.60	0.5638
SUBJECT(SEQUENCE)		004		1	1.5200	0.6395	8	2.38	0.0448
SUBJECT(SEQUENCE)		007		1	0	.	.	.	.
SUBJECT(SEQUENCE)		005		2	-0.5500	0.6395	8	-0.86	0.4148
SUBJECT(SEQUENCE)		006		2	0.5000	0.6395	8	0.78	0.4568
SUBJECT(SEQUENCE)		008		2	-1.0100	0.6395	8	-1.58	0.1529
SUBJECT(SEQUENCE)		009		2	-1.6400	0.6395	8	-2.56	0.0334
SUBJECT(SEQUENCE)		010		2	0	.	.	.	.

Type 3 Tests of Fixed Effects
Effect	Num DF	Den DF	F Value	Pr > F
TREATMENT	1	8	14.26	0.0054
PERIOD	1	8	0.16	0.6956
SEQUENCE	1	8	0.11	0.7457
SUBJECT(SEQUENCE)	8	8	3.12	0.0641

Least Squares Means
Effect	TREATMENT	Estimate	Standard Error	DF	t Value	Pr > \|t\|	Alpha	Lower	Upper
TREATMENT	A	3.7080	0.2022	8	18.34	<.0001	0.1	3.3320	4.0840
TREATMENT	B	2.6280	0.2022	8	13.00	<.0001	0.1	2.2520	3.0040

Differences of Least Squares Means
Effect	TREATMENT	_TREATMENT	Estimate	Standard Error	DF	t Value	Pr > \|t\|	Alpha	Lower	Upper
TREATMENT	B	A	-1.0800	0.2860	8	-3.78	0.0054	0.1	-1.6118	-0.5482

Re: when using proc mixed to treat subject as random effects or fixed effects, why get same result?

annaliseshen — Fri, 13 Jan 2017 22:19:15 GMT

continuing from my previous question, same dataset

Why when we use proc mixed procedure and treat subject as random or fixed effects,same answer got .

annaliseshen — Fri, 13 Jan 2017 22:22:59 GMT

PROC MIXED DATA=TRY;
CLASS TREATMENT PERIOD SEQUENCE SUBJECT;
MODEL CONC=TREATMENT PERIOD SEQUENCE/SOLUTION;
RANDOM SUBJECT(SEQUENCE);
LSMEANS TREATMENT/PDIFF=control("A","B") CL ALPHA=0.1 ;
RUN;

PROC MIXED DATA=TRY;
CLASS TREATMENT PERIOD SEQUENCE SUBJECT;
MODEL CONC=TREATMENT PERIOD SEQUENCE SUBJECT(SEQUENCE)/SOLUTION;
LSMEANS TREATMENT/PDIFF=control("A","B") CL ALPHA=0.1 ;
RUN;

The SAS System

If WE treat subject(sequence ) as fixed effect

The Mixed Procedure

Model Information
Data Set	WORK.TRY
Dependent Variable	CONC
Covariance Structure	Diagonal
Estimation Method	REML
Residual Variance Method	Profile
Fixed Effects SE Method	Model-Based
Degrees of Freedom Method	Residual

Class Level Information
Class	Levels	Values
TREATMENT	2	A B
PERIOD	2	1 2
SEQUENCE	2	1 2
SUBJECT	10	001 002 003 004 005 006 007 008 009 010

Dimensions
Covariance Parameters	1
Columns in X	17
Columns in Z	0
Subjects	1
Max Obs Per Subject	20

Number of Observations
Number of Observations Read	20
Number of Observations Used	20
Number of Observations Not Used	0

Covariance Parameter Estimates
Cov Parm	Estimate
Residual	0.4089

Fit Statistics
-2 Res Log Likelihood	25.7
AIC (smaller is better)	27.7
AICC (smaller is better)	28.4
BIC (smaller is better)	27.8

Solution for Fixed Effects
Effect	TREATMENT	SUBJECT	PERIOD	SEQUENCE	Estimate	Standard Error	DF	t Value	Pr > \|t\|
Intercept					3.1780	0.4953	8	6.42	0.0002
TREATMENT	A				1.0800	0.2860	8	3.78	0.0054
TREATMENT	B				0	.	.	.	.
PERIOD			1		-0.1160	0.2860	8	-0.41	0.6956
PERIOD			2		0	.	.	.	.
SEQUENCE				1	-0.7100	0.6395	8	-1.11	0.2991
SEQUENCE				2	0	.	.	.	.
SUBJECT(SEQUENCE)		001		1	-0.1750	0.6395	8	-0.27	0.7913
SUBJECT(SEQUENCE)		002		1	-0.4000	0.6395	8	-0.63	0.5490
SUBJECT(SEQUENCE)		003		1	0.3850	0.6395	8	0.60	0.5638
SUBJECT(SEQUENCE)		004		1	1.5200	0.6395	8	2.38	0.0448
SUBJECT(SEQUENCE)		007		1	0	.	.	.	.
SUBJECT(SEQUENCE)		005		2	-0.5500	0.6395	8	-0.86	0.4148
SUBJECT(SEQUENCE)		006		2	0.5000	0.6395	8	0.78	0.4568
SUBJECT(SEQUENCE)		008		2	-1.0100	0.6395	8	-1.58	0.1529
SUBJECT(SEQUENCE)		009		2	-1.6400	0.6395	8	-2.56	0.0334
SUBJECT(SEQUENCE)		010		2	0	.	.	.	.

Type 3 Tests of Fixed Effects
Effect	Num DF	Den DF	F Value	Pr > F
TREATMENT	1	8	14.26	0.0054
PERIOD	1	8	0.16	0.6956
SEQUENCE	1	8	0.11	0.7457
SUBJECT(SEQUENCE)	8	8	3.12	0.0641

Least Squares Means
Effect	TREATMENT	Estimate	Standard Error	DF	t Value	Pr > \|t\|	Alpha	Lower	Upper
TREATMENT	A	3.7080	0.2022	8	18.34	<.0001	0.1	3.3320	4.0840
TREATMENT	B	2.6280	0.2022	8	13.00	<.0001	0.1	2.2520	3.0040

Differences of Least Squares Means
Effect	TREATMENT	_TREATMENT	Estimate	Standard Error	DF	t Value	Pr > \|t\|	Alpha	Lower	Upper
TREATMENT	B	A	-1.0800	0.2860	8	-3.78	0.0054	0.1	-1.6118	-0.5482

Re: Why when we use proc mixed procedure and treat subject as random or fixed effects,same answer go

SteveDenham — Tue, 17 Jan 2017 14:34:29 GMT

See my answer to this here:

https://communities.sas.com/t5/SAS-Statistical-Procedures/When-random-effects-exists-why-are-the-results-same-from-proc/m-p/324701#U324701

Steve Denham