Table 14.1 on page 214, intercept only model. Since there is no within variance at level-1, there is no within statement in this model.
Title: multilevel logistic regression
Data:
File is ch14_1.dat ;
Variable:
Names are
respnr ltbm eltwm sex age edu relserv class reg cohab;
Missing are all (-9999) ;
usevar are cohab reg;
categorical is cohab;
cluster is reg;
Analysis:
Type = twolevel random ;
model:
%between%
cohab;
TESTS OF MODEL FIT
Loglikelihood
H0 Value -1420.493
Information Criteria
Number of Free Parameters 2
Akaike (AIC) 2844.986
Bayesian (BIC) 2856.266
Sample-Size Adjusted BIC 2849.912
(n* = (n + 2) / 24)
MODEL RESULTS
Estimates S.E. Est./S.E.
Within Level
Between Level
Thresholds
COHAB$1 0.278 0.062 4.442
Variances
COHAB 0.032 0.019 1.635
Table 14.2 on page 217, model 1.
Title: multilevel logistic regression
Data:
File is ch14_1.dat ;
Variable:
Names are
respnr ltbm eltwm sex age edu relserv class reg cohab;
Missing are all (-9999) ;
usevar are cohab reg agec agec2 age3 age4 ;
within are agec agec2 age3 age4 ;
categorical is cohab;
cluster is reg;
Define:
agec = age - 20;
agec2 = (age-20)**2;
IF (age>30) THEN age3 = (age-30)**2;
IF (age <=30) THEN age3 = 0;
IF (age > 40) THEN age4 = (age-40)**2;
IF (age <=40) THEN age4 = 0;
Analysis:
Type = twolevel random ;
model:
%within%
cohab on agec agec2 age3 age4 ;
%between%
cohab;
TESTS OF MODEL FIT
Loglikelihood
H0 Value -1123.849
Information Criteria
Number of Free Parameters 6
Akaike (AIC) 2259.697
Bayesian (BIC) 2293.535
Sample-Size Adjusted BIC 2274.473
(n* = (n + 2) / 24)
MODEL RESULTS
Estimates S.E. Est./S.E.
Within Level
COHAB ON
AGEC 0.545 0.044 12.408
AGEC2 -0.029 0.003 -8.800
AGE3 0.024 0.005 4.760
AGE4 0.007 0.002 3.150
Between Level
Thresholds
COHAB$1 1.214 0.157 7.755
Variances
COHAB 0.063 0.024 2.605
Table 14.2 on page 217, model 2.
Title: multilevel logistic regression
Data:
File is ch14_1.dat ;
Variable:
Names are
respnr ltbm eltwm sex age edu relserv class reg cohab;
Missing are all (-9999) ;
usevar are cohab reg agec agec2 age3 age4 religion;
within are agec agec2 age3 age4 religion;
categorical is cohab;
cluster is reg;
Define:
agec = age - 20;
agec2 = (age-20)**2;
IF (age>30) THEN age3 = (age-30)**2;
IF (age <=30) THEN age3 = 0;
IF (age > 40) THEN age4 = (age-40)**2;
IF (age <=40) THEN age4 = 0;
IF (relserv <3.5) THEN religion = 1;
if (relserv >=3.5) THEN religion = 0;
Analysis:
Type = twolevel random ;
model:
%within%
cohab on agec agec2 age3 age4 religion;
%between%
cohab;
TESTS OF MODEL FIT
Loglikelihood
H0 Value -1085.620
Information Criteria
Number of Free Parameters 7
Akaike (AIC) 2185.240
Bayesian (BIC) 2224.717
Sample-Size Adjusted BIC 2202.477
(n* = (n + 2) / 24)
MODEL RESULTS
Estimates S.E. Est./S.E.
Within Level
COHAB ON
AGEC 0.549 0.044 12.338
AGEC2 -0.029 0.003 -8.409
AGE3 0.024 0.005 4.388
AGE4 0.008 0.002 3.243
RELIGION -1.870 0.224 -8.344
Between Level
Thresholds
COHAB$1 1.117 0.154 7.262
Variances
COHAB 0.051 0.025 2.070