Table 3.
Linear regression model for activity-based exposure concentration
| Explanatory variable | Coefficienta (β i, γ i) | T statistic | Significance |
|---|---|---|---|
| Constantb (β0) | 11.5 | 35 | 3e-180 |
| Black | 1.2 | 2.9 | 3e-3 |
| Below poverty | 1.7 | 4.3 | 2e-5 |
| Middle incomec | 1.2 | 4.7 | 2e-6 |
| Urban | 8.3 | 23 | 1e-98 |
| Suburban | 3.4 | 11 | 8e-25 |
| Second city | 2.6 | 7.5 | 1e-13 |
| Time away from residence | 0.2 | 6.8 | 1e-11 |
| Goodness of fit | |||
| R 2 | 0.40 | ||
| Adjusted R 2 | 0.39 | ||
| Number of cases | 1,120 | ||
aThe regression model is C A (μg/m3) = β 0 + β 1 X c1 + β 2 X c2 + …γ 1 X t1 + γ 2 X t2 + … + ε, where X ci ∈ {1,0} are the binary variables and X ti are the continuous variables. Only the time variable was entered as a continuous variable (with units of hours). β i have units of μg/m3, γ i have units of (μg/m3)·h
bThe constant concentration represents exposures for people who are nonblack with incomes over $75,000, who live in town/rural areas and did not travel on the sample day
cMiddle income refers to households with income above the poverty threshold but less than $75,000