Sampling Techniques in SPC
Simple Random sampling provides the basis for several probability sampling methods. With simple random sampling, all possible sample of size 'n' has the probability of being selected.
Stratified random sampling is a variations of random sampling in which the population is partitioned into comparatively homogeneous groups called as strata & a simple random sample is chosen from each stratum. After that the results from the strata are aggregated to make inferences regarding the population. Another benefit of this method is that inferences regarding the subpopulation represented by each stratum might also be made.
It involves partitioning the population into separate groups called as clusters. Unlike the case of stratified sample random sampling, this is desirable for the clusters to be composed of heterogeneous units. In single-stage cluster sampling, the sampled clusters. In two-stage cluster sampling, a simple random sample of clusters is choosing and then a simple random sample is chosen from the units in each sampled cluster. One of the primary applications of cluster sampling is known as area sampling, where the clusters are city blocks, townships or other well-defined geographic sections of the population.
In several segments of industry, the application of Statistical Process Control (SPC) has become a common practice. This is more so in failure data analysis of equipment in the mass production industries such like automobile, automobile component or consumer electronics manufacture. SPC is broadly applied within the organizations for machine process capability & machine efficiency estimations.
The application of procedure control charts is only one element of utilizes of SPC. Many of the benefits of using a systematic & statistical approach to analyzing why a specific process varies are derived well before control charts are generated. Indeed very frequently improved process understanding leading to enhanced process capability and therefore indicates the machine efficiency and also might reduce or even eliminate the need for process control chart.