|
Modele 7 |
The LOGISTIC Procedure
|
Model Information |
|
|---|---|
|
Data Set |
DOC.DATA4 |
|
Response Variable |
sgr |
|
Number of Response Levels |
6 |
|
Model |
cumulative logit |
|
Optimization Technique |
Fisher's scoring |
|
Number of Observations Read |
470 |
|---|---|
|
Number of Observations Used |
470 |
|
Response Profile |
||
|---|---|---|
|
Ordered |
sgr |
Total |
|
1 |
5 |
7 |
|
2 |
4 |
31 |
|
3 |
3 |
74 |
|
4 |
2 |
106 |
|
5 |
1 |
140 |
|
6 |
0 |
112 |
|
Probabilities modeled are cumulated over the lower Ordered Values. |
|
Class Level Information |
||
|---|---|---|
|
Class |
Value |
Design |
|
ansmedit |
Non |
0 |
|
|
Oui |
1 |
|
ansm |
0 |
0 |
|
|
1 |
1 |
|
faculte2 |
NonPa |
1 |
|
|
Paris |
-1 |
|
Model Convergence Status |
|---|
|
Convergence criterion (GCONV=1E-8) satisfied. |
|
Score
Test for the Proportional |
||
|---|---|---|
|
Chi-Square |
DF |
Pr > ChiSq |
|
8.7071 |
16 |
0.9250 |
|
Model Fit Statistics |
||
|---|---|---|
|
Criterion |
Intercept Only |
Intercept
and |
|
AIC |
1487.164 |
1458.136 |
|
SC |
1507.928 |
1495.510 |
|
-2 Log L |
1477.164 |
1440.136 |
|
Testing Global Null Hypothesis: BETA=0 |
|||
|---|---|---|---|
|
Test |
Chi-Square |
DF |
Pr > ChiSq |
|
Likelihood Ratio |
37.0286 |
4 |
<.0001 |
|
Score |
34.8379 |
4 |
<.0001 |
|
Wald |
36.3968 |
4 |
<.0001 |
|
Type 3 Analysis of Effects |
|||
|---|---|---|---|
|
Effect |
DF |
Wald |
Pr > ChiSq |
|
age |
1 |
14.3011 |
0.0002 |
|
ansmedit |
1 |
4.9418 |
0.0262 |
|
faculte2 |
1 |
6.4789 |
0.0109 |
|
ansm |
1 |
5.1727 |
0.0229 |
|
Analysis of Maximum Likelihood Estimates |
||||||
|---|---|---|---|---|---|---|
|
Parameter |
|
DF |
Estimate |
Standard |
Wald |
Pr > ChiSq |
|
Intercept |
5 |
1 |
-5.4731 |
0.4793 |
130.4117 |
<.0001 |
|
Intercept |
4 |
1 |
-3.6761 |
0.3311 |
123.2552 |
<.0001 |
|
Intercept |
3 |
1 |
-2.3466 |
0.2965 |
62.6503 |
<.0001 |
|
Intercept |
2 |
1 |
-1.2754 |
0.2816 |
20.5077 |
<.0001 |
|
Intercept |
1 |
1 |
0.0939 |
0.2765 |
0.1154 |
0.7341 |
|
age |
|
1 |
0.0235 |
0.00622 |
14.3011 |
0.0002 |
|
ansmedit |
Oui |
1 |
0.4816 |
0.2167 |
4.9418 |
0.0262 |
|
faculte2 |
NonPa |
1 |
0.2134 |
0.0838 |
6.4789 |
0.0109 |
|
ansm |
1 |
1 |
0.8162 |
0.3589 |
5.1727 |
0.0229 |
|
Odds Ratio Estimates |
|||
|---|---|---|---|
|
Effect |
Point Estimate |
95%
Wald |
|
|
age |
1.024 |
1.011 |
1.036 |
|
ansmedit Oui vs Non |
1.619 |
1.059 |
2.475 |
|
faculte2 NonPa vs Paris |
1.532 |
1.103 |
2.128 |
|
ansm 1 vs 0 |
2.262 |
1.119 |
4.570 |
|
Association
of Predicted Probabilities and |
|||
|---|---|---|---|
|
Percent Concordant |
60.1 |
Somers' D |
0.220 |
|
Percent Discordant |
38.1 |
Gamma |
0.224 |
|
Percent Tied |
1.8 |
Tau-a |
0.171 |
|
Pairs |
85517 |
c |
0.610 |