Stratified random sampling approach

Figure 2.3 Stratified random sampling. This approach divides the site into smaller zones. Samples are collected randomly within each zone, and could even be composited.

The most common sampling modes are SRS, discussed previously, stratified random sampling, and systematic random sampling. SRS over time or space consists of identifying times or places totally at random to take the samples. Hie great disadvantage of this approach is that certain portions of the lot or production times may be under or over represented, and process stability cannot be monitored effectively. Consequently, we do not recommend SRS for long-term examination of lot characteristics. 

Figure 2.6 Stratified systematic sampling or systematic nonaligned sampling. This approach combines random and systematic sampling. The site is divided into regular spaced regions, and then a sample unit is chosen randomly within each of these regions.

Stratified sampling is commonly used in a location which is known to have contaminants heterogeneously distributed. This is therefore the most common approach to sampling. In this type of sampling, the site is sub-divided into smaller areas, each of which is fairly homogeneous, and thus more accurate sampling can take place. Each sub-area is then randomly sampled. The sub-dividing of the... 

Characteristics of composite sampling are that the samples (or measurements) are taken following a random approach and then bulked. The best use is for relatively homogeneous material, unless stratified composites are created. A phase I ESA investigation often uses composite samples. 

---