Data Sampling Techniques - Meaning and Different Types of Techniques
The method which is adopted to collect the sample obviously has large implications on the conclusions drawn from that sample. Mentioned below are the techniques which can be used for sampling populations as well as processes. A brief detail has been given so that you can understand the pros, cons and correct usages of each of these techniques:
Population Sampling Techniques
- Random Sampling: A random sample is a sample in which every member of a population has an equal chance of being selected. As one can understand from the definition this method is not applicable to the results of processes because the population set should be static. In case of a process, the population set is dynamic and new and new results are constantly added to the data. Therefore one cannot ensure that every member has an equal chance of being selected. However the results of random sampling are amongst the best if adequate sample size is selected.
- Stratified Random Sampling: In case of stratified random sampling, the population is broken down into strata which contain their own data elements. Within the strata, each data element has an equal chance of being selected. However the number of elements from each starta are pre-determined. This is close to random sampling. However, once again it cannot be used for a process because it requires a static population whereas a process is dynamic by definition.
Process Sampling Techniques
- Systematic Sampling: In case of systematic sampling, the first element in the sample is chosen at random. Then the next elements are chosen in a systematic fashion. For example, the first element will be chosen at random then every tenth element will be included in the sample. Since these types of samples are systematic and do not need a static population base, they can be used for process sampling. In fact systematic sampling is one of the most popular methods used for process sampling.
- Rational Subgrouping: Rational subgrouping is a sampling technique whose main aim is to produce data for control charts. Samples are drawn from subgroups at regular intervals. Hence the person who is collecting the sample needs to decide the sample size as well as the interval. This should be large enough to detect any changes in the underlying process.
The advantage of this strategy is that you have an additional dimension which is time. Studies can be done and cycles and patterns can be found out. The disadvantage is that planning one of these studies requires considerably more expertise than a random sampling plan.
Although there are many more types of sampling techniques that are available, the four listed above account for almost 80% of the types of samples that are used in studies. Hence a thorough understanding of these four will help improve your sampling endeavour.
|❮ Previous Article||Next Article ❯|
About the Author(s)
MSG team comprises experienced faculty and professionals who develop the content for the portal. We collectively refer to our team as - MSG Experts. To Know more, click on About Us.
- Introduction to Measure Phase
- Outputs in the Measure Phase
- What is a Detailed Process Map ?
- How to Create a Detailed Process Map ?
- Identify the Vital Few Inputs
- Characteristics of Data
- Different types of Data
- Data Shapes & Characteristics of Shapes
- Data Collection Plan
- Data Sampling Techniques
- Understanding Measurement Error
- Importance of Measurement Systems Analysis
- Causes of Measurement Variation
- Accuracy vs. Precision
- Linearity and Resolution
- Steps Involved in Conducting a Measurement System Analysis