Schemea |
n
|
n
b
g
|
CV
|
\(\bar{n}\)
|
---|
1 | 3, 4, 5 | 889, 667, 533 | 0.250 | 4 |
2 | 3, 5 | 1335, 799 | 0.354 | 4 |
3 | 2, 4, 6 | 1333, 666, 445 | 0.500 | 4 |
4 | 2, 6 | 2002, 666 | 0.707 | 4 |
5 | 5, 6, 7 | 534, 445, 380 | 0.167 | 6 |
6 | 5, 7* | 799, 572* | 0.236 | 6 |
7 | 4, 6, 8 | 666, 444, 334 | 0.333 | 6 |
8 | 4, 8 | 1000, 500 | 0.471 | 6 |
9 | 2, 6, 10 | 1333, 444, 267 | 0.667 | 6 |
10 | 2, 10 | 2000, 400 | 0.943 | 6 |
11 | 7, 8, 9** | 380, 333, 297** | 0.125 | 8 |
12 | 7, 9 | 572, 444 | 0.177 | 8 |
13 | 6, 8, 10 | 446, 333, 266 | 0.250 | 8 |
14
|
6, 10***
|
666, 400***
|
0.353
|
8
|
15 | 4, 8, 12 | 666, 334, 222 | 0.500 | 8 |
16
|
4, 12
|
1001, 333
|
0.707
|
8
|
17 | 2, 8, 14 | 1334, 334, 190 | 0.750 | 8 |
18
|
2, 14
|
2005, 285
|
1.010
|
8
|
- For some schemes, a few individuals were removed if they could not make up a group for the group sizes of that scheme
- For the schemes in italics (14, 16 and 18), the two-family designs were also tested
- n group size, ng number of groups, CV coefficient of variation and \(\bar{n}\) average group size
- *One individual was removed; **three individuals were removed; ***four individuals were removed
- aFor each scheme, three dilution parameters were evaluated
- bThe 8000 individuals were divided into either two or three subsets depending on the number of group sizes in a scheme and the subsets were divided by the group size, resulting in the number of groups