Statistical
modelling/Milk offtake
The output also shows how care must be taken in interpreting
the parameter estimates. Each is corrected for the others in the model with the
t-value measuring the significance of the parameter when included in addition to
all other parameters in the model.
The accumulated analysis of variance, on the
other hand, shows the additional sum of squares accounted for as each variable
is added in turn. The order in which the terms are included to the model is
important. Each sum of squares is corrected for variables already included in
the model but not for those to be added later. Therefore the F-value has a
different interpretation from the t-value.
|
|
Response variate: TOTALM
Fitted terms: Constant+LOCATION+AGEC+AGEC.LOCATION
*** Summary of analysis ***
|
d.f. |
s.s. |
m.s. |
v.r. |
Regression |
3 |
30.23 |
10.0766 |
31.39 |
Residual |
160 |
51.37 |
0.3210 |
|
Total |
163 |
81.60 |
0.5006 |
|
Percentage variance accounted for 35.9
*** Estimates of parameters ***
|
estimate |
s.e. |
t(160) |
Constant |
0.983 |
0.161 |
6.09 |
LOCATION 1 |
1.411 |
0.212 |
6.67 |
AGEC |
0.0073 |
0.0250 |
0.29 |
AGEC.LOCATION 1 |
-0.1036 |
0.0333 |
-3.11
|
Parameters for factors are differences compared
with the reference level: Factor Reference level: LOCATION 2
*** Accumulated analysis of variance ***
Change |
d.f. |
s.s. |
m.s. |
v.r. |
+ LOCATION |
1 |
24.0448 |
24.0448 |
74.90 |
+ AGEC |
1 |
3.0777 |
3.0777 |
9.59 |
+ AGEC.LOCATION |
1 |
3.1073 |
3.1073 |
9.68 |
Residual |
160 |
51.3656 |
0.3210 |
|
Total |
163 |
81.5953 |
0.5006 |
|
|
|