3.2.2. Résultats des estimations empiriques
Le modèle logit multinomial estimé sous le logiciel
Stata version 8, permet d'obtenir les résultats consignés dans le
tableau ci-dessous :
Tableau4 : Résultats de l'estimation du
modèle logit multinomial
Variable dépendante : tech (nombre de technologies CES)
|
Variables
|
Coefficients
|
Std. Err
|
Z
|
Prob
|
Effet
|
Z
|
Prob
|
|
explicatives
|
|
|
|
|
marginal
|
|
|
|
P (tech=1) cout
|
-6.311008***
|
1.146958
|
-5.50
|
0.000
|
-0.0345488
|
-1.21
|
0.227
|
|
facilte
|
0.1244665
|
0.2675408
|
0.47
|
0.642
|
0.0028201
|
0.63
|
0.529
|
|
utilite
|
-0.1963432
|
0.3173703
|
-0.62
|
0.536
|
-0.0088051
|
-1.11
|
0.266
|
|
superficie
|
0.5927408
|
0.384076
|
1.54
|
0.123
|
-0.0004451
|
-0.08
|
0.934
|
|
KINI Janvier, 12eme Promotion,
DEA/PTCI
|
|
52
|
|
|
|
|
|
|
|
|
|
taille
|
-0.1903279*
|
0.1127115
|
-1.69
|
0.091
|
-0.0026911
|
-1.00
|
0.315
|
|
escompt
|
-0.4874255
|
0.9535067
|
-0.51
|
0.609
|
0.0046512
|
0.29
|
0.770
|
|
risque
|
-0.7076787
|
0.6751335
|
-1.05
|
0.295
|
-0.0112062
|
-0.93
|
0.351
|
|
cons
|
92.83205
|
17.28966
|
5.37
|
0.000
|
-
|
-
|
-
|
|
P (tech=2) cout
|
-5.084995***
|
1.052767
|
-4.83
|
0.000
|
-0.3640332
|
-3.52
|
0.000
|
|
facilte
|
0.0057664
|
0.1857118
|
0.03
|
0.975
|
0.0172843
|
0.48
|
0.633
|
|
utilite
|
0.2291505
|
0.1758194
|
1.30
|
0.192
|
-0.017326
|
-0.38
|
0.704
|
|
superficie
|
0.7373844***
|
0.2795553
|
2.64
|
0.008
|
0.0813171
|
1.80
|
0.072
|
|
taille
|
-0.0709134
|
0.0618854
|
-1.15
|
0.252
|
-0.0123922
|
-0.95
|
0.340
|
|
escompt
|
-1.1359***
|
0.4612785
|
-2.46
|
0.014
|
-0.2736788
|
-2.42
|
0.016
|
|
risque
|
-0.1386203
|
0.4685017
|
-0.30
|
0.767
|
|
-0.04
|
0.967
|
|
|
|
|
|
-0.0036721
|
|
|
|
cons
|
77.5068
|
15.44663
|
5.02
|
0.000
|
-
|
-
|
-
|
|
P (tech=3) cout
|
6.311007***
|
1.146958
|
5.50
|
0.000
|
0.1466089
|
1.55
|
0.122
|
|
facilte
|
-0.1244665
|
0.2675408
|
0.47
|
0.642
|
0.00065
|
0.13
|
0.899
|
|
utilite
|
0.1963432
|
0.3173702
|
0.62
|
0.536
|
-0.0082387
|
-1.00
|
0.316
|
|
superficie
|
-0.5927408
|
0.384076
|
-1.54
|
0.123
|
-0.0198798
|
-1.49
|
0.135
|
|
taille
|
0.1903279*
|
0.1127115
|
1.69
|
0.091
|
0.0016906
|
0.86
|
0.392
|
|
escompt
|
0.4874255
|
0.9535067
|
0.51
|
0.609
|
0.0234432
|
1.04
|
0.299
|
|
risque
|
0.7076787
|
0.6751335
|
1.05
|
0.295
|
0.0042991
|
0.32
|
0.749
|
|
cons
|
-92.83205
|
17.28966
|
-5.37
|
0.000
|
-
|
-
|
-
|
|
P (tech=4) cout
|
-3.64195***
|
0.9988913
|
-3.65
|
0.000
|
0.2519731
|
2.69
|
0.007
|
|
facilte
|
-0.0940674
|
0.1803052
|
-0.52
|
0.602
|
-0.0207544
|
-0.59
|
0.557
|
|
utilite
|
0.3775534*
|
0.2277186
|
1.66
|
0.097
|
0.0343698
|
0.75
|
0.451
|
|
superficie
|
0.3982703
|
0.2571337
|
1.55
|
0.121
|
-0.0609923
|
-1.39
|
0.165
|
|
taille
|
-0.0046353
|
0.0534778
|
-0.09
|
0.931
|
0.0133927
|
1.08
|
0.279
|
|
escompt
|
0.1490242
|
0.5230435
|
0.28
|
0.776
|
0.2455845
|
2.17
|
0.030
|
|
risque
|
-0.0954299
|
0.4104068
|
-0.23
|
0.816
|
0.0105791
|
0.13
|
0.900
|
|
KINI Janvier, 12eme Promotion,
DEA/PTCI
|
|
53
|
|
|
|
|
|
|
|
|
|
cons
|
47.5002
|
14.65516
|
3.24
|
0.001
|
-
|
-
|
-
|
|
P (tech=5) cout
|
-4.045883***
|
1.374478
|
-2.94
|
0.003
|
4.38e-10
|
0.00
|
1.000
|
|
Facilite
|
0.1033059
|
0.5123983
|
0.20
|
0.840
|
1.09e-10
|
-
|
-
|
|
Utilité
|
5.540855
|
1469.605
|
0.00
|
0.997
|
4.69e-09
|
0.00
|
0.999
|
|
superficie
|
0.7181979
|
0.6465843
|
1.11
|
0.267
|
9.10e-11
|
-
|
-
|
|
Taille
|
-0.4074238
|
0.3237387
|
-1.26
|
0.208
|
-3.15e-10
|
-0.05
|
0.958
|
|
escompt
|
0.6032348
|
1.35169
|
0.45
|
0.655
|
1.18e-09
|
0.00
|
0.999
|
|
Risque
|
16.50057
|
4408.814
|
0.00
|
0.997
|
1.48e-08
|
0.00
|
0.999
|
|
Cons
|
-46.24354
|
-
|
-
|
-
|
-
|
-
|
-
|
|
Mac Fadden
|
R2 = 0,5448
|
|
|
|
|
|
|
Sources : Données de l'enquête
réalisée par l'auteur Z : Statistique test de la
loi normale
Prob : probabilité de
significativité des coefficients du model
| 
