Example C3 Binary response model

Raudenbush and Bhumirat (1992) analysed data on children repeating a grade during their time at primary school. The data were from a national survey of primary education in Thailand in 1988, we use a sub set of that data here.

Reference

Raudenbush, S.W., Bhumirat, C., 1992. The distribution of resources for primary education and its consequences for educational achievement in Thailand, International Journal of Educational Research, 17, 143-164

Data description

Number of observations (rows): 7185

Number of variables (columns): 4

The variables include the following:

schoolid =school identifier

sex= 1 if child is male, 0 otherwise

pped=1 if the child had pre primary experience, 0 otherwise

repeat=1 if the child repeated a grade during primary school, 0 otherwise

The first few lines of thaieduc1.dat look like

Sabre commands

out thaieduc.log

trace thaieduc.trace

data schoolid sex pped repeat

case schoolid

yvar repeat

constant cons

lfit cons

dis m

dis e

fit cons

dis m

dis e

data schoolid sex pped repeat msesc

case schoolid

yvar repeat

constant cons

lfit msesc sex pped cons

dis m

dis e

fit msesc sex pped cons

dis m

dis e

stop

Sabre log file

<S> trace thaieduc.trace

<S> data schoolid sex pped repeat

8582 observations in dataset

<S> case schoolid

<S> yvar repeat

<S> constant cons

<S> lfit cons

Iteration       Log. lik.       Difference

__________________________________________

1          -5948.5891

2          -3625.9614        2323.

3          -3554.2465        71.71

4          -3553.4908       0.7557

5          -3553.4906       0.1283E-03

6          -3553.4906       0.1296E-09

<S> dis m

X-vars            Y-var

______________________________

cons              repeat

Univariate model

Standard logit

Number of observations             =    8582

X-var df           =     1

Log likelihood =     -3553.4906     on    8581 residual degrees of freedom

<S> dis e

Parameter              Estimate         Std. Err.

___________________________________________________

cons                   -1.7738          0.30651E-01

<S> fit cons

Initial Homogeneous Fit:

Iteration       Log. lik.       Difference

__________________________________________

1          -5948.5891

2          -3625.9614        2323.

3          -3554.2465        71.71

4          -3553.4908       0.7557

5          -3553.4906       0.1283E-03

6          -3553.4906       0.1296E-09

Iteration       Log. lik.         Step      End-points     Orthogonality

length    0          1      criterion

________________________________________________________________________

1          -3237.7286        1.0000    fixed  fixed       334.98

2          -3219.2799        1.0000    fixed  fixed       202.12

3          -3217.3207        1.0000    fixed  fixed       167.22

4          -3217.2649        1.0000    fixed  fixed       167.84

5          -3217.2642        1.0000    fixed  fixed       216.57

6          -3217.2642        1.0000    fixed  fixed       148.29

7          -3217.2642        1.0000    fixed  fixed

<S> dis m

X-vars            Y-var             Case-var

________________________________________________

cons              repeat            schoolid

Univariate model

Standard logit

Gaussian random effects

Number of observations             =    8582

Number of cases                    =     411

X-var df           =     1

Scale df           =     1

Log likelihood =     -3217.2642     on    8580 residual degrees of freedom

<S> dis e

Parameter              Estimate         Std. Err.

___________________________________________________

cons                   -2.1263          0.79655E-01

scale                   1.2984          0.84165E-01

<S> data schoolid sex pped repeat msesc

--- new analysis begins

7516 observations in dataset

<S> case schoolid

<S> yvar repeat

<S> constant cons

<S> lfit msesc sex pped cons

Iteration       Log. lik.       Difference

__________________________________________

1          -5209.6942

2          -3094.0634        2116.

3          -3007.1924        86.87

4          -3004.7357        2.457

5          -3004.7313       0.4417E-02

6          -3004.7313       0.1669E-07

<S> dis m

X-vars            Y-var

______________________________

cons              repeat

msesc

sex

pped

Univariate model

Standard logit

Number of observations             =    7516

X-var df           =     4

Log likelihood =     -3004.7313     on    7512 residual degrees of freedom

<S> dis e

Parameter              Estimate         Std. Err.

___________________________________________________

cons                   -1.7832          0.58777E-01

msesc                 -0.24149          0.93750E-01

sex                    0.42777          0.67637E-01

pped                  -0.56885          0.70421E-01

<S> fit msesc sex pped cons

Initial Homogeneous Fit:

Iteration       Log. lik.       Difference

__________________________________________

1          -5209.6942

2          -3094.0634        2116.

3          -3007.1924        86.87

4          -3004.7357        2.457

5          -3004.7313       0.4417E-02

6          -3004.7313       0.1669E-07

Iteration       Log. lik.         Step      End-points     Orthogonality

length    0          1      criterion

________________________________________________________________________

1          -2741.1866        1.0000    fixed  fixed       193.41

2          -2723.0976        1.0000    fixed  fixed       178.19

3          -2720.8948        1.0000    fixed  fixed       89.392

4          -2720.7670        1.0000    fixed  fixed       40.916

5          -2720.7589        1.0000    fixed  fixed       33.123

6          -2720.7582        1.0000    fixed  fixed       23.011

7          -2720.7581        1.0000    fixed  fixed

<S> dis m

X-vars            Y-var             Case-var

________________________________________________

cons              repeat            schoolid

msesc

sex

pped

Univariate model

Standard logit

Gaussian random effects

Number of observations             =    7516

Number of cases                    =     356

X-var df           =     4

Scale df           =     1

Log likelihood =     -2720.7581     on    7511 residual degrees of freedom

<S> dis e

Parameter              Estimate         Std. Err.

___________________________________________________

cons                   -2.2280          0.10461

msesc                 -0.41370          0.22462

sex                    0.53177          0.75805E-01

pped                  -0.64022          0.98885E-01

scale                   1.3026          0.72601E-01

<S> stop