Statistical modelling
We can fit the model by using the dialog box shown below obtained by
Stats → Regression Analysis → Generalized Linear Models... Then by clicking the Options button,
then ticking Accumulated, an analysis of
variance is shown which gives the sums of squares accounted for by each parameter in the model.
The interpretation of the AGEWEAN term is simpler than
that for a factor with discrete levels such as DAMAGE7. This is a continuous
covariate and so the value of 0.07058 (± 0.00886) kg/day represents the slope
(± s.e.) of the linear regression of WEANWT on AGEWEAN adjusted for all other
factors.
|
|
***** Regression Analysis *****
Response variate: WEANWT
Fitted terms: Constant
+ GENOTYPE + YEAR +
SEX + AGEWEAN + DAMAGE7
*** Estimates of parameters
***
Parameter
|
estimate
|
s.e.
|
t
|
Constant
|
4.327
|
0.883
|
4.90
|
GENOTYPE DR
|
-0.493
|
0.306
|
-1.61
|
GENOTYPE RD
|
-0.408
|
0.222
|
-1.84
|
GENOTYPE RR
|
-1.008
|
0.272
|
-3.71
|
YEAR 92
|
-1.551
|
0.308
|
-5.03
|
YEAR 93
|
-1.228
|
0.291
|
-4.22
|
YEAR 94
|
-2.983
|
0.388
|
-7.69
|
YEAR 95
|
-3.258
|
0.346
|
-9.40
|
YEAR 96
|
-2.333
|
0.423
|
-5.51
|
SEX M
|
0.482
|
0.170
|
2.84
|
AGEWEAN
|
0.07058
|
0.00886
|
7.97
|
DAMAGE7 3
|
1.833
|
0.319
|
5.75
|
DAMAGE7 4
|
2.741
|
0.331
|
8.28
|
DAMAGE7 5
|
2.742
|
0.322
|
8.52
|
DAMAGE7 6
|
2.322
|
0.382
|
6.07
|
DAMAGE7 7
|
1.754
|
0.462
|
3.79
|
DAMAGE7 >=8
|
1.405
|
0.647
|
2.17
|
*** Accumulated analysis of variance ***
Change |
d.f. |
s.s. |
m.s. |
v.r. |
+GENOTYPE |
3 |
570.427 |
190.142 |
38.68 |
+YEAR |
5 |
735.646 |
147.129 |
29.93 |
+SEX |
1 |
59.013 |
59.013 |
12.00 |
+AGEWEAN |
1 |
336.792 |
336.792 |
68.51 |
+DAMAGE7 |
6 |
445.076 |
74.179 |
15.09 |
Residual |
683 |
3357.495 |
4.916 |
|
|
|
|
|
|
Total |
699 |
5504.450 |
7.875 |
|
|
|