Data Collection of Y project (Rational Subgrouping).

To conduct data collection, first we must ensure that our measurement system to collect data is properly appropriate by using GR&R Study. Then, we can collect the data.

To collect Data, we must have the way how to collect it. We cannot just take the data without know how our process is. Generally, our product/process has a variation as long as its process describe in the picture below.

Picture M.4 Rational Subgrouping

In the sequences of process, we can select the data. The possible way to collect the data:

- Selecting n sampling data which are randomly …… (1)

Picture M.5 Individual Random Sampling

- Selecting n sampling data which are sequentially ……… (2)

Picture M.6 Group Sampling

- Selecting number of group, which n sampling data selected sequentially. Group should be selected by randomly as many as possible. …….. (3)

Picture M.7 Rational Subgrouping

In the case of (1) and (2), we cannot detect and separate the within variation and between variation. Within variation is variation which caused by internal causes / common causes that uncontrollable. Within variation can be derived by measurement variation which allowed by study. This type of variation called White Noise. Between variation is the variation which caused by external causes (5M1E), assignable factors / special causes and could be controlled. Between variation often called Black Noise (See picture M.4!). Because of disable to separate within and between variation, so this type of collecting data will not allow to estimate the short-term and long-term variation. Short-term is the condition that represent if all the special causes removed from the process. While the Long-term is the condition that represent involves all causes that affect the product/process.

In the case of (3) we can separate the within variation and between variation. Therefore, we can estimate the short-term and long-term variation. This type of Data collection called Rational Subgrouping.