Geoscience Reference
In-Depth Information
0
2
4
6810
89
90
0.5
0.4
0.3
0.2
0.1
85
86
87
88
0.5
0.4
0.3
0.2
0.1
81
82
83
84
0.5
0.4
0.3
0.2
0.1
0
2
4
6810
0
Ye ar
2
4
6
810
FIGURE 11.3
Scatter plots of mercury concentrations for 12 years of the study for each of the stations (sta-
tions are numbered 81-90).
The least-squares fits of the linear regression models for each site are easily
obtained using most software packages. Once the regression models are esti-
mated for each station, a display of confidence intervals for the intercepts and
slopes associated with the sampling unit can help in making sensible conclu-
sions about the individual units as well as general statements about the popu-
lation of sampling units. Figure 11.4 shows 95% confidence intervals for the
intercepts and slopes (denoted by “year”) for each of the 10 stations in the study.
In each panel of Figure 11.4, a vertical line segment extends from zero on
the horizontal axis. For every station with the exception of station 82, the
mercury concentrations were low at the start of the study period because
in the left-hand panel of Figure 11.4 the vertical line represents zero mer-
cury concentration at the start of the study. It appears as if stations 84, 87,
and 90 began the study with the lowest levels of contamination, and sta-
tion 82 began the study with the highest level. In the right-hand panel of
Figure 11.4, stations whose corresponding confidence intervals lie to the
