Table 6.
Linear regression analysis of stationary area measurements with inhalable cobalt as the independent variable and other particle fractions as dependenta,b
| Aerosol fraction (Dependent; Independent) | Regression | Prediction | |||
|---|---|---|---|---|---|
| β (p) | 95% CI (β) | r 2 | Predicted concentrationc | 95% CI (y)d | |
|
PM2.5Co (S);
Inhalable Co (S) |
0.345(0.002) | 0.139–0.552 | 0.332 | 0.0000726 | 0.0000112–0.0000495 |
|
PM10Co (S);
Inhalable Co (S) |
0.831(<0.001) | 0.591–1.07 | 0.680 | 0.00557 | 0.000599–0.0518 |
|
Particle surface area (mean) (S);
Inhalable Co (S) |
−0.201(0.037) | −0.387 to (−0.014) | 0.221 | 2.89 | 0.770–10.8 |
|
Particle number (mean) (S);
Inhalable Co (S) |
−0.040(0.509) | −0.164 to 0.084 | 0.019 | 2276 | 742–6970 |
β, regression coefficient; r 2, correlation coefficient; S, stationary area measurements; 95% CI (β), 95% confidence interval for β.
aAll data ln-transformed, except predicted concentrations presented in original units.
b n = 28, except for particle surface area (n = 22) and particle number (n = 27).
cPredicted concentration = predicted individual concentration for inhalable cobalt 0.02mg m−3.
dCI (y) = 95% confidence interval for predicted dependent metrics, individual concentrations.