Information Technology Reference
In-Depth Information
1.293 which indicates a positive auto-correlation, nevertheless the dimension of the
sample is big enough to assume an asymptotic independence. Table 3 shows the
values of the multiple linear regression model. The VIF values are all lower than 5,
which do not indicate problems in terms of multicollinearity. It is also possible to
identify by the model the variables that are significant predictors for the Quality of
Life: years of smoking (p value 0.049) and size (p value < 0.001).
Fig. 2. Normal distribution and homogeneity of errors
Table 3. Multiple Linear Regression model
Variables
B
P value
VIF
Educational Level
-0.110
0.091
1.238
Marital Status
-0.170
0.274
1.220
Years of Smoking
-0.013
0.049
2.428
Number of cigarettes per day
0.002
0.902
2.131
Years of drinking
-0.038
0.959
1.927
Liters of beer per day
-0.139
0.705
2.217
Liters of hard alcohol per day
-0.338
0.443
2.471
Size
-0.158
< 0.001
2.048
Local metastasis
-0.257
0.172
1.755
Metastasis distance
0.043
0.382
1.420
Histopathological diagnosis
-0.528
0.309
1.781
Tracheostomy
-0.170
0.067
3.363
Supply Type
-0.721
0.063
3.476
Liters of wine per day
0.208
0.057
2.818
Smoking
0.021
0.955
1.344
Years of ex-smoker
-0.042
0.560
1.501
Moment
0.038
0.201
1.257
Voice prosthesis
0.619
0.095
1.725
Constant
3.940
< 0.001
--
 
Search WWH ::




Custom Search