Table 3. Optimal cluster allocation as a function of sample size.
| (1) Number of Clusters Selected | (2) Minimum Number of Records Per Sample | (3) nh[h] = sample size in clusters (sections) for stratum = h | (4) Nh[h] = total number of clusters (sections) for stratum = h | (5) bh[h] = upper boundary for stratum[h]. | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| – | – | h = 1 | h = 2 | h = 3 | h = 4 | h = 1 | h = 2 | h = 3 | h = 4 | h = 1 | h = 2 | h = 3 | h = 4 |
| 4 | 373 | 1 | 1 | 1 | 1 | 6 | 5 | 6 | 3 | 51.0 | 110.5 | 175.5 | 208.0 |
| 5 | 454 | 1 | 1 | 2 | 1 | 5 | 4 | 8 | 3 | 40.0 | 84.0 | 175.5 | 208.0 |
| 6 | 570 | 1 | 1 | 3 | 1 | 5 | 4 | 8 | 3 | 40.0 | 84.0 | 175.5 | 208.0 |
| 7 | 697 | 1 | 1 | 4 | 1 | 5 | 4 | 8 | 3 | 40.0 | 84.0 | 175.5 | 208.0 |
| 8 | 714 | 2 | 1 | 4 | 1 | 5 | 4 | 8 | 3 | 40.0 | 84.0 | 175.5 | 208.0 |
| 9 | 850 | 2 | 1 | 5 | 1 | 5 | 4 | 8 | 3 | 40.0 | 84.0 | 175.5 | 208.0 |
| 10 | 986 | 2 | 1 | 6 | 1 | 5 | 4 | 8 | 3 | 40.0 | 84.0 | 175.5 | 208.0 |
| 11 | 1138 | 2 | 1 | 7 | 1 | 5 | 4 | 8 | 3 | 40.0 | 84.0 | 175.5 | 208.0 |
| 12 | 1164 | 3 | 1 | 7 | 1 | 5 | 4 | 8 | 3 | 40.0 | 84.0 | 175.5 | 208.0 |
| 13 | 1780 | 1 | 1 | 1 | 12 | 2 | 3 | 3 | 12 | 21.5 | 40.0 | 62.5 | 208.0 |
| 14 | 1839 | 1 | 1 | 1 | 13 | 2 | 3 | 2 | 13 | 21.5 | 40.0 | 57.5 | 208.0 |
| 15 | 1883 | 1 | 1 | 1 | 14 | 2 | 2 | 2 | 14 | 21.5 | 33.5 | 51.0 | 208.0 |
| 16 | 1895 | 1 | 1 | 13 | 2 | 2 | 3 | 13 | 2 | 21.5 | 40.0 | 201.0 | 208.0 |
| 17 | 1929 | 1 | 1 | 6 | 10 | 2 | 2 | 6 | 10 | 21.5 | 33.5 | 103.5 | 208.0 |
| 18 | 1886 | 2 | 2 | 2 | 12 | 2 | 3 | 3 | 12 | 21.5 | 40.0 | 62.5 | 208.0 |
| 19 | 1946 | 2 | 1 | 6 | 10 | 2 | 2 | 6 | 10 | 21.5 | 33.5 | 103.5 | 208.0 |
| 20 | 1979 | 2 | 2 | 2 | 14 | 2 | 2 | 2 | 14 | 21.5 | 33.5 | 51.0 | 208.0 |
Stratification variables for selecting clusters (sections) for one-stage stratified cluster sampling with 4 levels of stratification (L = 4). The entries in each column are (1) the number of clusters to be selected (2) the minimum number of PSU’s (i.e. residential structures) spanned by the selected clusters if the allocation nh[h] is drawn (3) the number of clusters (sections) nh[h] to be drawn from each stratum (4) the total number of clusters Nh[h] in each stratum and (5) the upper boundary bh[h] in units of “residential structures per cluster” for each of the four strata. A comparable table was constructed for stratification by population per section, but is not shown for the sake of brevity.