VI. ANNEXES
ANNEXE I. Tableau des résultats de la
régression initiale
Regress previdu idueffectif incidrate meanage overdose chomage
pibhabppa valeur_idh traitement clients centre pes,r
Regression with robust standard errors Number of obs
= 50
F( 11, 38)
= 19.14
Prob > F
= 0.0000
R-squared
= 0.5080
Root MSE
= 21.331
---------------------------------------------------------------------------
| Robust
previdu | Coef. Std. Err. t P>|t| [95%
Conf. Interval]
-------------+-------------------------------------------------------------
idueffectif | .0000111 5.43e-06 2.05 0.047
1.31e-07 0000221
incidrate | 1.201233 .4606312 2.61 0.013
.2687337 2.133732
meanage | -.0257832 .8705752 -0.03 0.977
-1.78817 1.736604
overdose | .0056711 .0095543 0.59 0.556
-.0136706 .0250127
chomage | -.1595139 .5879769 -0.27 0.788
-1.349811 1.030783
pibhabppa | .0000994 .0003841 0.26 0.797
-.0006782 .0008769
valeur_idh | -173.1131 73.51714 -2.35 0.024
-321.9408 24.28543
traitement | -.0031496 .0014054 -2.24 0.031
-.0059947 .0003045
clients | .0004263 .0000801 5.32 0.000
.0002641 .0005884
centre | .0241874 .0114346 2.12 0.041
.0010393 .0473355
pes | -3.87e-06 1.21e-06 -3.20 0.003
-6.32e-06 1.42e-06
_cons | 163.8215 66.60221 2.46 0.019
28.9924 298.6507
---------------------------------------------------------------------------
ANNEXE II. Test de normalité
Regression with robust standard errors Number of obs
= 50
F( 11, 38)
= 19.14
Prob > F
= 0.0000
R-squared
= 0.5080
Root MSE
= 21.331
---------------------------------------------------------------------------
| Robust
previdu | Coef. Std. Err. t P>|t| [95%
Conf. Interval]
-------------+-------------------------------------------------------------
idueffectif | .0000111 5.43e-06 2.05 0.047
1.31e-07 .0000221
incidrate | 1.201233 .4606312 2.61 0.013
.2687337 2.133732
meanage | -.0257832 .8705752 -0.03 0.977
-1.78817 1.736604
overdose | .0056711 .0095543 0.59 0.556
-.0136706 .0250127
chomage | -.1595139 .5879769 -0.27 0.788
-1.349811 1.030783
pibhabppa | .0000994 .0003841 0.26 0.797
-.0006782 .0008769
valeur_idh | -173.1131 73.51714 -2.35 0.024
-321.9408 24.28543
traitement | -.0031496 .0014054 -2.24 0.031
-.0059947 .0003045
clients | .0004263 .0000801 5.32 0.000
.0002641 .0005884
centre | .0241874 .0114346 2.12 0.041
.0010393 .0473355
pes | -3.87e-06 1.21e-06 -3.20 0.003
-6.32e-06 1.42e-06
_cons | 163.8215 66.60221 2.46 0.019
28.9924 298.6507
---------------------------------------------------------------------------
Residuals
-------------------------------------------------------------
Percentiles Smallest
1% -31.57689 -31.57689
5% -25.5694 -29.72784
10% -23.73396 -25.5694 Obs 50
25% -12.12354 -24.08225 Sum of Wgt. 50
50% -4.075961 Mean -1.37e-08
Largest Std. Dev. 18.78496
75% 11.42489 33.88633
90% 25.83068 40.23976 Variance 352.8747
95% 40.23976 41.40499 Skewness
.6315624
99% 47.37689 47.37689 Kurtosis
2.929241
ANNEXE III. Le test
d'hétéroscédasticité
Regress previdu idueffectif incidrate meanage overdose chomage
pibhabppa valeur_idh traitement clients centre pes
. predict resbp, resid
. generate resbp2=resbp^2
Regress resbp2 idueffectif incidrate meanage overdose chomage
pibhabppa valeur_idh traitement clients centre pes
Source | SS df MS Number of obs
= 50
-------------+------------------------------ F( 11, 38)
= 2.12
Model | 4383996.98 11 398545.18 Prob > F
= 0.0430
Residual | 7151854.66 38 188206.702 R-squared
= 0.3800
-------------+------------------------------ Adj R-squared
= 0.2006
Total | 11535851.6 49 235425.544 Root MSE
= 433.83
---------------------------------------------------------------------------
resbp2 | Coef. Std. Err. t P>|t| [95%
Conf. Interval]
-------------+-------------------------------------------------------------
idueffectif | -.0000351 .0001575 -0.22 0.825
-.0003539 .0002837
incidrate | 3.935181 8.291757 0.47 0.638
-12.8506 20.72097
meanage | 11.14967 16.66308 0.67 0.507
-22.58297 44.88231
overdose | -.2653693 .4473032 -0.59 0.557
-1.170887 .6401487
chomage | -12.22165 12.58458 -0.97 0.338
-37.6978 13.25449
pibhabppa | -.0096393 .007913 -1.22 0.231
-.0256585 .0063798
valeur_idh | -2308.746 1187.322 -1.94 0.059
-4712.353 94.86097
traitement | .027644 .0598373 0.46 0.647
-.0934902 .1487782
clients | -.001334 .0036402 -0.37 0.716
-.0087032 .0060353
centre | -.0238316 .5391229 -0.04 0.965
-1.115229 1.067566
pes | .0000249 .0000489 0.51 0.614
-.0000741 .0001239
_cons | 2305.629 1075.128 2.14 0.038
129.1453 4482.112
---------------------------------------------------------------------------
ANNEXE IV. Test
d'endogénéité
Regress valeur_idh tx_brut_de_scolarisation idueffectif incidrate
meanage overdose chomage pibhabppa traitement clients centre pes,r
Regression with robust standard errors Number of
obs =50
F( 11,
38) =11.43
Prob >
F = 0.0000
R-squared
= 0.7065
Root MSE =
0.05334
---------------------------------------------------------------------------
| Robust
valeur_idh | Coef. Std. Err. t P>|t| [95%
Conf. Interval]
-------------+-------------------------------------------------------------
tx_brut_de~n | .002418 .000827 2.92 0.006
.0007437 .0040922
idueffectif | 4.46e-08 9.17e-09 4.86 0.000
2.60e-08 6.31e-08
incidrate | -.0008902 .0005294 -1.68 0.101
-.0019619 .0001815
meanage | -.0008009 .0018737 -0.43 0.671
-.0045941 .0029923
overdose | -.0000294 .0000309 -0.95 0.348
-.0000919 .0000332
chomage | -.0009041 .0016853 -0.54 0.595
-.0043158 .0025076
pibhabppa | 3.31e-06 1.02e-06 3.25 0.002
1.25e-06 5.37e-06
traitement | -6.78e-07 4.00e-06 -0.17 0.866
-8.77e-06 7.42e-06
clients | -1.05e-07 2.39e-07 -0.44 0.663
-5.90e-07 3.80e-07
centre | .0000155 .0000357 0.43 0.667
-.0000568 .0000878
pes | 1.47e-09 3.19e-09 0.46 0.648
-4.99e-09 7.93e-09
_cons | .6255717 .0703957 8.89 0.000
.4830631 .7680803
---------------------------------------------------------------------------
. predict resinstr,resid
Regress previdu idueffectif incidrate meanage overdose chomage
pibhabppa valeur_idh traitement clients centre pes resinstr,r
Regression with robust standard errors Number of obs
= 50
F( 12, 37)
= 17.45
Prob > F
= 0.0000
R-squared
= 0.5132
Root MSE
= 21.503
---------------------------------------------------------------------------
| Robust
previdu | Coef. Std. Err. t P>|t| [95%
Conf. Interval]
-------------+-------------------------------------------------------------
idueffectif | 7.13e-06 6.31e-06 1.13 0.265
-5.64e-06 .0000199
incidrate | 1.214726 .4570347 2.66 0.012
.2886857 2.140766
meanage | .0317762 .8494158 0.04 0.970
-1.689304 1.752856
overdose | .0094068 .0107482 0.88 0.387
-.0123712 .0311848
chomage | -.0446534 .6671568 -0.07 0.947
-1.396441 1.307135
pibhabppa | -.0002331 .0004868 -0.48 0.635
-.0012193 .0007532
valeur_idh | -96.76065 116.4929 -0.83 0.412
-332.7978 139.2765
traitement | -.0034893 .001481 -2.36 0.024
-.0064901 .0004885
clients | .0004683 .0000937 5.00 0.000
.0002785 .0006582
centre | .0200383 .0118306 1.69 0.099
-.0039327 .0440093
pes | -4.33e-06 1.36e-06 -3.17 0.003
-7.09e-06-1.56e-06
resinstr | -94.2836 170.4148 -0.55 0.583
-439.5767 251.0095
_cons | 102.1134 94.5675 1.08 0.287
-89.49854 293.7254
ANNEXE V. Test de Chow
Source | SS df MS Number of obs
= 50
-------------+------------------------------ F( 11, 38)
= 3.57
Model | 17850.8105 11 1622.80096 Prob > F
= 0.0017
Residual | 17290.8603 38 455.022638 R-squared
= 0.5080
-------------+------------------------------ Adj R-squared
= 0.3655
Total | 35141.6708 49 717.176955 Root MSE
= 21.331
---------------------------------------------------------------------------
previdu | Coef. Std. Err. t P>|t| [95%
Conf. Interval]
-------------+-------------------------------------------------------------
idueffectif | .0000111 7.74e-06 1.44 0.159
-4.56e-06 .0000268
incidrate | 1.201233 .4077045 2.95 0.005
.3758781 2.026588
meanage | -.0257832 .8193212 -0.03 0.975
-1.684412 1.632846
overdose | .0056711 .0219938 0.26 0.798
-.0388531 .0501953
chomage | -.1595139 .6187819 -0.26 0.798
-1.412172 1.093144
pibhabppa | .0000994 .0003891 0.26 0.800
-.0006883 .000887
valeur_idh | -173.1131 58.38044 -2.97 0.005
-291.2981-54.92807
traitement | -.0031496 .0029422 -1.07 0.291
-.0091058 .0028065
clients | .0004263 .000179 2.38 0.022
.0000639 .0007886
centre | .0241874 .0265086 0.91 0.367
-.0294765 .0778512
pes | -3.87e-06 2.41e-06 -1.61 0.116
-8.74e-06 1.00e-06
_cons | 163.8215 52.86391 3.10 0.004
56.80415 270.8389
Regress previdu idueffectif incidrate meanage overdose chomage
pibhabppa valeur_idh traitement clients centre pes if pibhabppa>18000
Source | SS df MS Number of obs
= 22
-------------+------------------------------ F( 11, 10)
= 9.85
Model | 8887.34153 11 807.940139 Prob > F
= 0.0006
Residual | 820.291705 10 82.0291705 R-squared
= 0.9155
-------------+------------------------------ Adj R-squared
= 0.8226
Total | 9707.63323 21 462.268249 Root MSE
= 9.057
---------------------------------------------------------------------------
previdu | Coef. Std. Err. t P>|t| [95%
Conf. Interval]
-------------+-------------------------------------------------------------
idueffectif | -.0000102 .0000136 -0.75 0.470
-.0000404 .00002
incidrate | 4.094068 1.012481 4.04 0.002
1.838119 6.350016
meanage | 1.125665 1.255657 0.90 0.391
-1.672112 3.923443
overdose | .0011097 .0124186 0.09 0.931
-.0265606 .0287801
chomage | 1.48155 1.411707 1.05 0.319
-1.66393 4.62703
pibhabppa | .0001381 .0003228 0.43 0.678
-.0005811 .0008573
valeur_idh | -114.5285 122.3645 -0.94 0.371
-387.1737 158.1167
traitement | -.0025899 .0022812 -1.14 0.283
-.0076728 .0024929
clients | .0004796 .0000998 4.81 0.001
.0002574 .0007019
centre | .0285544 .0186532 1.53 0.157
-.0130075 .0701164
pes | -3.84e-06 1.25e-06 -3.08 0.012
-6.61e-06-1.06e-06
_cons | 56.72072 112.6078 0.50 0.625
-194.1851 307.6265
---------------------------------------------------------------------------
ANNEXE VI. Test de Ramsey Reset
Source | SS df MS Number of obs
= 50
-------------+------------------------------ F( 11, 38)
= 3.57
Model | 17850.8105 11 1622.80096 Prob > F
= 0.0017
Residual | 17290.8603 38 455.022638 R-squared
= 0.5080
-------------+------------------------------ Adj R-squared
= 0.3655
Total | 35141.6708 49 717.176955 Root MSE
= 21.331
---------------------------------------------------------------------------
previdu | Coef. Std. Err. t P>|t| [95%
Conf. interval]
-------------+-------------------------------------------------------------
idueffectif | .0000111 7.74e-06 1.44 0.159
-4.56e-06 .0000268
incidrate | 1.201233 .4077045 2.95 0.005
.3758781 2.026588
meanage | -.0257832 .8193212 -0.03 0.975
-1.684412 1.632846
overdose | .0056711 .0219938 0.26 0.798
-.0388531 .0501953
chomage | -.1595139 .6187819 -0.26 0.798
-1.412172 1.093144
pibhabppa | .0000994 .0003891 0.26 0.800
-.0006883 .000887
valeur_idh | -173.1131 58.38044 -2.97 0.005
-291.2981 54.92807
traitement | -.0031496 .0029422 -1.07 0.291
-.0091058 .0028065
clients | .0004263 .000179 2.38 0.022
.0000639 .0007886
centre | .0241874 .0265086 0.91 0.367
-.0294765 .0778512
pes | -3.87e-06 2.41e-06 -1.61 0.116
-8.74e-06 1.00e-06
_cons | 163.8215 52.86391 3.10 0.004
56.80415 270.8389
---------------------------------------------------------------------------
. predict previduest
(option xb assumed; fitted values)
. g previduest2= previduest^2
. g previduest3= previduest^3
. g previduest4= previduest^4
Regress previdu idueffectif incidrate meanage overdose chomage
pibhabppa valeur_idh traitement clients centre pes previduest2
previduest3previduest4
Source | SS df MS Number of obs
= 50
-------------+------------------------------ F( 14, 35)
= 6.01
Model | 24815.8618 14 1772.56156 Prob > F
= 0.0000
Residual | 10325.809 35 295.023114 R-squared
= 0.7062
-------------+------------------------------ Adj R-squared
= 0.5886
Total | 35141.6708 49 717.176955 Root MSE
= 17.176
---------------------------------------------------------------------------
previdu | Coef. Std. Err. t P>|t| [95%
Conf. Interval]
-------------+-------------------------------------------------------------
idueffectif | -.0000268 .0000176 -1.52 0.137
-.0000625 8.93e-06
incidrate | 4.790416 2.149717 2.23 0.032
.4262576 9.154574
meanage | .2726361 .6878789 0.40 0.694
-1.123832 1.669105
overdose | .0015743 .022081 0.07 0.944
-.0432524 .046401
chomage | .4855452 .5657022 0.86 0.397
-.6628913 1.633982
pibhabppa | -.0002491 .0003297 -0.76 0.455
-.0009185 .0004203
valeur_idh | -52.96368 249.2496 -0.21 0.833
-558.9672 453.0399
traitement | -.0028901 .0047903 -0.60 0.550
-.012615 .0068348
clients | .0005808 .000626 0.93 0.360
-.0006901 .0018517
centre | .0038469 .0419211 0.09 0.927
-.0812576 .0889513
pes | -1.86e-06 5.64e-06 -0.33 0.744
-.0000133 9.60e-06
previduest2 | -.0699248 .0842013 -0.83 0.412
-.2408626 .101013
previduest3 | .003237 .0018172 1.78 0.084
-.0004522 .0069261
previduest4 | -.0000314 .0000123 -2.54 0.016
-.0000564-6.30e-06
_cons | 48.01249 231.2964 0.21 0.837
-421.5442 517.5692
---------------------------------------------------------------------------
ANNEXE VI: autre modèle
Regress incidrate previdu idupercent meanage traitement clients
pes centre overdose chomage pibhabppa valeur_idh traitementcarre,r
Regression with robust standard errors Number of
obs = 50
F( 12,
37) = 4.28
Prob >
F 0.0003
R-squared
0.3378
Root MSE
7.9239
---------------------------------------------------------------------------
| Robust
incidrate | Coef. Std. Err. t P>|t|
[95%ConfInterval]
-------------+-------------------------------------------------------------
previdu | .186336 .0897284 2.08 0.045
.0045291.368143
idupercent | -.3601886 .6187677 -0.58 0.564
-1.613931.8935539
meanage | -.2066356 .2507545 -0.82 0.415
-.7147125 .3014413
traitement | -.0012841 .0007177 -1.79 0.082
-.0027383 .00017
clients | -.0000783 .0000427 -1.83 0.075
-.0001649 8.23e-06
pes | 6.78e-07 6.24e-07 1.09 0.284
-5.86e-07 1.94e-06
centre | -.0087584 .0032708 -2.68 0.011
-.0153856 -0021311
overdose | .0021907 .0044452 0.49 0.625
-.0068162 111975
chomage | -.0483714 .1225859 -0.39 0.695
-.2967539 .2000112
pibhabppa | 9.04e-06 .0000726 0.12 0.901
-.000138 0001561
valeur_idh | 44.41831 21.21018 2.09 0.043
1.44241 7.39421
traitement~e | 1.59e-07 7.46e-08 2.13 0.040
7.77e-09 3.10e-07
_cons | -30.59444 14.37229 -2.13 0.040
-59.71546-1.473414
--------------------------------------------------------------------------
. sum res,detail
Residuals
-------------------------------------------------------------
Percentiles Smallest
1% -9.64874 -9.64874
5% -7.24662 -7.538783
10% -6.289858 -7.24662 Obs 50
25% -2.76622 -6.990764 Sum of Wgt. 50
50% -.4493213 Mean -2.50e-08
Largest Std. Dev. 6.885615
75% 1.651354 3.698311
90% 3.249649 5.016016 Variance 47.41169
95% 5.016016 5.620093 Skewness
4.43267
99% 41.72096 41.72096 Kurtosis
28.29055
| 
