Table 4

 Contrast, bias, precision, and Spearman correlation between the least square mean estimates for the different groups based on the 1988 and 1997 exposure survey

Grouping variableContrastComparison of groups*
ε 97ε 88No. of groupsBias§Precision¶Correlation Spearman
*Comparison of groups was limited to data from the seven plants that were represented in both surveys.
†Contrast between groups based on exposure data from the 1997 exposure survey (n = 590; k = 220).
‡Contrast between groups based on exposure data from the 1988 exposure survey (n = 591; k = 244).
§Mean difference between the 1988 and 1997 estimates in log-space of the individual group estimates; the exponent of the calculated bias (exp(Ygi,88 − Ygi,97) provides the ratio between the 1997 and 1988 estimates on the original scale.
¶Standard deviation in the difference in individual group estimates based on the 1988 and 1997 survey.
**Determinant based grouping is based on exposure determinants identified in the 1997 survey.
Inhalable dust
Department0.020.0670.470.220.29
Plant0.050.0070.440.260.39
Type0.000.0030.490.180.50
Determinant based**0.280.1130.530.260.50
Plant/department0.160.27380.530.970.53
Plant/determinant based0.490.19200.500.610.47
Department/determinant based0.400.18190.480.420.62
Dermal CSM
Department0.060.1870.670.320.71
Plant0.030.0170.610.33−0.29
Type0.000.0030.760.120.50
Determinant based**0.340.1430.740.261.00
Plant/department0.250.32380.640.930.55
Plant/determinant based0.320.20200.670.760.45
Department/determinant based0.280.20190.670.470.64