Logistic Regression

 [DataSet3] 

Case Processing Summary

Unweighted Casesa

N

Percent

Selected Cases

Included in Analysis

30

100.0

Missing Cases

0

.0

Total

30

100.0

Unselected Cases

0

.0

Total

30

100.0

a. If weight is in effect, see classification table for the total number of cases.

 

Dependent Variable Encoding

Original Value

Internal Value

No Violation

0

Violation

1

 

Categorical Variables Codings

 

Frequency

Parameter coding

(1)

(2)

(3)

(4)

prior

None

6

1.000

.000

.000

.000

Fine/Probation

6

.000

1.000

.000

.000

Reform School

6

.000

.000

1.000

.000

Jail

6

.000

.000

.000

1.000

Penitentiary

6

.000

.000

.000

.000

prsnpnsh

None

10

1.000

.000

 

 

1-2 Times

10

.000

1.000

 

 

3 or More

10

.000

.000

 

 

 


Block 0: Beginning Block

 

Classification Tablea,b

 

Observed

Predicted

 

prlvlt

Percentage Correct

 

No Violation

Violation

Step 0

prlvlt

No Violation

2167

0

100.0

Violation

796

0

.0

Overall Percentage

 

 

73.1

a. Constant is included in the model.

 

 

b. The cut value is .500

 

 

 

 

Variables in the Equation

 

B

S.E.

Wald

df

Sig.

Exp(B)

Step 0

Constant

-1.001

.041

583.905

1

.000

.367

 

Variables not in the Equation

 

Score

df

Sig.

Step 0

Variables

prsnpnsh

53.879

2

.000

prsnpnsh(1)

53.513

1

.000

prsnpnsh(2)

22.877

1

.000

prior

87.198

4

.000

prior(1)

69.575

1

.000

prior(2)

.003

1

.955

prior(3)

36.753

1

.000

prior(4)

13.866

1

.000

Overall Statistics

119.873

6

.000

 


Block 1: Method = Enter

 

Omnibus Tests of Model Coefficients

 

Chi-square

df

Sig.

Step 1

Step

116.396

6

.000

Block

116.396

6

.000

Model

116.396

6

.000

 

Model Summary

Step

-2 Log likelihood

Cox & Snell R Square

Nagelkerke R Square

1

3331.992a

.039

.056

a. Estimation terminated at iteration number 4 because parameter estimates changed by less than .001.

 

Classification Tablea

 

Observed

Predicted

 

prlvlt

Percentage Correct

 

No Violation

Violation

Step 1

prlvlt

No Violation

2167

0

100.0

Violation

796

0

.0

Overall Percentage

 

 

73.1

a. The cut value is .500

 

 

 

 

Variables in the Equation

 

B

S.E.

Wald

df

Sig.

Exp(B)

Step 1a

prsnpnsh

 

 

33.057

2

.000

 

prsnpnsh(1)

-.502

.119

17.787

1

.000

.605

prsnpnsh(2)

-.010

.131

.005

1

.942

.990

prior

 

 

64.476

4

.000

 

prior(1)

-.825

.172

22.971

1

.000

.438

prior(2)

-.606

.214

7.978

1

.005

.546

prior(3)

-.012

.197

.004

1

.952

.988

prior(4)

-.225

.199

1.276

1

.259

.798

Constant

-.127

.182

.488

1

.485

.880

a. Variable(s) entered on step 1: prsnpnsh, prior.

 

 

 

 


Block 2: Method = Enter

 

Omnibus Tests of Model Coefficients

 

Chi-square

df

Sig.

Step 1

Step

6.089

8

.637

Block

6.089

8

.637

Model

122.485

14

.000

 

Model Summary

Step

-2 Log likelihood

Cox & Snell R Square

Nagelkerke R Square

1

3325.902a

.040

.059

a. Estimation terminated at iteration number 4 because parameter estimates changed by less than .001.

 

Classification Tablea

 

Observed

Predicted

 

prlvlt

Percentage Correct

 

No Violation

Violation

Step 1

prlvlt

No Violation

2122

45

97.9

Violation

742

54

6.8

Overall Percentage

 

 

73.4

a. The cut value is .500

 

 

 

 

Variables in the Equation

 

B

S.E.

Wald

df

Sig.

Exp(B)

Step 1a

prsnpnsh

 

 

1.820

2

.403

 

prsnpnsh(1)

-.492

.400

1.513

1

.219

.611

prsnpnsh(2)

-.113

.432

.069

1

.793

.893

prior

 

 

8.116

4

.087

 

prior(1)

-.785

.352

4.965

1

.026

.456

prior(2)

-.460

.445

1.064

1

.302

.632

prior(3)

-.165

.392

.176

1

.675

.848

prior(4)

-.411

.409

1.011

1

.315

.663

prior * prsnpnsh

 

 

6.042

8

.642

 

prior(1) by prsnpnsh(1)

-.052

.435

.015

1

.904

.949

prior(1) by prsnpnsh(2)

-.063

.473

.018

1

.894

.939

prior(2) by prsnpnsh(1)

-.253

.546

.215

1

.643

.776

prior(2) by prsnpnsh(2)

-.064

.593

.012

1

.913

.938

prior(3) by prsnpnsh(1)

-.054

.494

.012

1

.912

.947

prior(3) by prsnpnsh(2)

.561

.530

1.119

1

.290

1.752

prior(4) by prsnpnsh(1)

.207

.505

.168

1

.682

1.230

prior(4) by prsnpnsh(2)

.277

.543

.261

1

.609

1.320

Constant

-.100

.317

.100

1

.752

.905

a. Variable(s) entered on step 1: prior * prsnpnsh .

 

 

 

 

 

 



[DataSet3] 

 

 

 



[DataSet3] 


Generalized Linear Models

 

 



[DataSet3] 

Model Information

Dependent Variable

prlvlta

Probability Distribution

Binomial

Link Function

Logit

a. The procedure models Violation as the response, treating No Violation as the reference category.

 

Case Processing Summary

 

N

Percent

Unweighted N

Included

2963

100.0%

30

Excluded

0

.0%

0

Total

2963

100.0%

30

 

Categorical Variable Information

 

N

Percent

Unweighted N

Dependent Variable

prlvlt

No Violation

2167

73.1%

15

Violation

796

26.9%

15

Total

2963

100.0%

30

Factor

prsnpnsh

None

1827

61.7%

10

1-2 Times

710

24.0%

10

3 or More

426

14.4%

10

Total

2963

100.0%

30

prior

None

1883

63.6%

6

Fine/Probation

262

8.8%

6

Reform School

331

11.2%

6

Jail

327

11.0%

6

Penitentiary

160

5.4%

6

Total

2963

100.0%

30

 

Goodness of Fitb

 

Value

df

Value/df

Deviance

6.089

8

.761

Scaled Deviance

6.089

8

 

Pearson Chi-Square

6.073

8

.759

Scaled Pearson Chi-Square

6.073

8

 

Log Likelihooda

-1.666E3

 

 

Akaike's Information Criterion (AIC)

3.346E3

 

 

Finite Sample Corrected AIC (AICC)

3.346E3

 

 

Bayesian Information Criterion (BIC)

3.388E3

 

 

Consistent AIC (CAIC)

3.395E3

 

 

Dependent Variable: prlvlt
Model: (Intercept), prsnpnsh, prior

 

a. The full log likelihood function is displayed and used in computing information criteria.

b. Information criteria are in small-is-better form.

 

Omnibus Testa

Likelihood Ratio Chi-Square

df

Sig.

116.396

6

.000

Dependent Variable: prlvlt
Model: (Intercept), prsnpnsh, prior

a. Compares the fitted model against the intercept-only model.

 

Tests of Model Effects

Source

Type III

Wald Chi-Square

df

Sig.

(Intercept)

121.359

1

.000

prsnpnsh

33.057

2

.000

prior

64.476

4

.000

Dependent Variable: prlvlt
Model: (Intercept), prsnpnsh, prior

 

 

Parameter Estimates

Parameter

B

Std. Error

95% Wald Confidence Interval

Hypothesis Test

Lower

Upper

Wald Chi-Square

df

Sig.

(Intercept)

-.127

.1824

-.485

.230

.488

1

.485

[prsnpnsh=0]

-.502

.1190

-.735

-.269

17.787

1

.000

[prsnpnsh=1]

-.010

.1308

-.266

.247

.005

1

.942

[prsnpnsh=2]

0a

.

.

.

.

.

.

[prior=0]

-.825

.1722

-1.163

-.488

22.971

1

.000

[prior=1]

-.606

.2144

-1.026

-.185

7.978

1

.005

[prior=2]

-.012

.1972

-.398

.375

.004

1

.952

[prior=3]

-.225

.1994

-.616

.166

1.276

1

.259

[prior=4]

0a

.

.

.

.

.

.

(Scale)

1b

 

 

 

 

 

 

Dependent Variable: prlvlt
Model: (Intercept), prsnpnsh, prior

 

 

 

 

a. Set to zero because this parameter is redundant.

 

 

 

b. Fixed at the displayed value.