Statistical modelling
An alternative approach to the statistical analysis is to
try the same model again but with the age of dam grouped into fewer discrete
categories. DAMAGE7 has been recoded as DAMAGE4 with four instead of seven age
categories. The four levels chosen for the new categories are, as suggested
earlier, namely 2, 3-4, 5-6 and 7-8 years.
This analysis gives a slightly higher residual mean square
than that originally obtained, i.e. an increase from 4.916 to 4.991
kg2.
|
|
***** Regression Analysis *****
Response variate: WEANWT
Fitted terms: Constant + GENOTYPE + YEAR +
SEX + AGEWEAN + DAMAGE4
*** Estimates of parameters ***
|
Parameter |
s.e. |
t(686) |
Constant |
4.170 |
0.887 |
4.70 |
GENOTYPE RD |
-0.297 |
0.221 |
-1.34 |
GENOTYPE RR |
-1.022 |
0.270 |
-3.79 |
GENOTYPE DR |
-0.573 |
0.302 |
-1.90 |
YEAR 92 |
-1.606 |
0.296 |
-5.42 |
YEAR 93 |
-1.363 |
0.281 |
-4.85 |
YEAR 94 |
-2.734 |
0.363 |
-7.54 |
YEAR 95 |
-3.155 |
0.345 |
-9.14 |
YEAR 96 |
-2.415 |
0.378 |
-6.39 |
SEX M |
0.493 |
0.171 |
2.89 |
AGEWEAN |
0.07242 |
0.00890 |
8.13 |
DAMAGE4 3-4 |
2.260 |
0.297 |
7.62 |
DAMAGE4 5-6 |
2.549 |
0.305 |
8.36 |
DAMAGE4 >=7 |
1.642 |
0.425 |
3.86 |
*** Accumulated analysis of variance ***
Change
|
d.f.
|
s.s.
|
m.s.
|
v.r.
|
+ GENOTYPE |
3 |
570.427 |
190.142 |
38.10 |
+ YEAR |
5 |
735.646 |
147.129 |
29.48 |
+ SEX |
1 |
59.013 |
59.013 |
11.82 |
+ AGEWEAN |
1 |
336.792 |
336.792 |
67.48 |
+ DAMAGE4 |
3 |
378.859 |
126.286 |
25.30 |
Residual |
686 |
3423.713 |
4.991 |
|
|
|
|
|
|
Total |
699 |
5504.450 |
7.875 |
|
|
|