There
are several poverty variables in the census data, but these individual
data elements have little to do with an appropriate and direct
deduction and modelling of poverty in a census tract. While recent
immigrant information is directly available, poverty variables such as
income level, education status, government transfer payments or
population without income are not given in relation to recent immigrant
but for the entire population. A statistical approach to analyse a
couple these information directly will be subjective rather than
objective and appropriate to revealing the truth in the data. An
approach adopted was to convert such information into percentages in
relation to their respective total samples. In this fashion, any sample
entity taken reflects its status relative to its parent sample space.
Subsequent to the above, a least squares model was determined for the
percent of recent immigrants with the chosen socio-economic variables.
These socio-economic variables considered are; low income economic
families, 15 years and more persons without income, 15 years and more
persons with education below grade 9 by highest level of schooling and
finally, government transfer payments. A summary of descriptive
regression model parameters and model performance measures calculated
for the percent of recent immigrants and the socio-economic variables
selected are shown below.
Percent of RIP and Percent of Low Income Economic Families
|
Montreal CMA
|
Montreal CMA |
Toronto CMA
|
Toronto CMA
|
Vancouver CMA
|
Vancouver CMA
|
|
1996 (N=769) |
2001 (N=930) |
1996 (N=813) |
2001 (N=985) |
1996 (N=299) |
2001 (N=424) |
| Regression
Equation |
1.73x+19.13 |
1.77x+12.7 |
1.22x+6.61 |
1.03x+5.64 |
0.87x+9.89 |
1.15x+7.40 |
| Coefficient of
correlation |
0.33 |
0.37 |
0.57
|
0.522 |
0.36 |
0.49 |
| Coefficient of
determination |
10.69
|
3.47 |
32.38 |
27.25 |
13.03 |
22.85 |
| Mean Bias Error
|
-0.03 |
-0.23 |
-0.094 |
-0.063 |
-0.07 |
-0.04 |
| Root Mean Square
Error |
13.09 |
9.98 |
7.64 |
7.071 |
7.79 |
6.97 |
| Mean Absolute
Error |
10.32 |
7.35 |
5.65 |
5.034 |
5.64 |
5.03 |
| Index of
Agreement |
0.70 |
0.73 |
0.86 |
0.824 |
0.96 |
0.80 |
Percent of RIP and Percent 15 Years & More Population Without Income
|
Montreal CMA |
Montreal CMA |
Toronto CMA |
Toronto CMA |
Vancouver CMA |
Vancouver CMA |
|
1996 (N=769) |
2001 (N=930) |
1996 (N=813) |
2001 (N=985) |
1996 (N=299) |
2001 (N=424) |
| Regression Equation |
0.05x+7.7 |
0.12x+4.79 |
0.27x+5.4 |
0.19x+4.04 |
0.23x+5.08 |
0.20x+4.26 |
| Coefficient of correlation |
0.01 |
0.07 |
0.46 |
0.369 |
0.36 |
0.29 |
| Coefficient of determination |
0.01 |
0.52 |
21.34 |
13.62 |
12.89 |
8.53 |
| Mean Bias Error |
-0.15 |
-0.09 |
-0.0718 |
-0.045 |
-0.03 |
-0.02 |
| Root Mean Square Error |
2.69 |
1.96 |
2.32 |
1.83 |
2.06 |
1.84 |
| Mean Absolute Error |
2.24 |
1.55 |
1.83 |
1.41 |
1.65 |
1.42 |
| Index of Agreement |
0.24 |
0.42 |
0.80 |
0.74 |
0.98 |
0.66 |
Percent of RIP and Government
Transfer Payments (%)