Exploration & description
The number of lambs characterised by age of dam also
reveals a frequency imbalance. The oldest Red Maasai ewes were aged 6 years
whereas one Dorper ewe was as old as 10 years. From the numbers of lambs for
each age category it can be seen that dams between the ages of 2 and 6 years
were most common. Extreme age classes of 1, 9 and 10 years had only one lamb
each. Since age of dam is a factor to be considered in the analysis of weaning
weight of lambs, it would not be sensible to keep these classes separate. One
can either omit these three records or pool them with existing ones. We have
chosen to put age 1 year and 2 years together to form one class (2 years and
below) and to put ages 9 and 10 years together into the age 8 year category to
form an '8 years and above' class. The column DAMAGE7 in CS3Data has been
created using the GenStat command Spread → Factor → Change levels… to put the extreme values into the neighbouring
categories.
DAMAGE |
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
Count |
GENOTYPE |
|
|
|
|
|
|
|
|
|
|
|
DD |
0 |
24 |
49 |
38 |
47 |
32 |
22 |
8 |
0 |
0 |
220 |
DR |
0 |
15 |
40 |
21 |
14 |
11 |
0 |
0 |
0 |
0 |
101 |
RD |
0 |
16 |
28 |
47 |
61 |
19 |
19 |
6 |
1 |
1 |
198 |
RR |
1 |
17 |
41 |
51 |
40 |
31 |
0 |
0 |
0 |
0 |
181 |
Count |
1 |
72 |
158 |
157 |
162 |
93 |
41 |
14 |
1 |
1 |
700 |
When fitting a classification factor in a statistical
model it is always important to check that there are reasonable numbers of
observations within each category level; attempts to fit parameter terms to
sparse data often leads to spurious estimates.
|