Multistage sampling, Sampling & Non-sampling Error || Bcis Notes

Multistage sampling:

Under the study when sampling is done in more than one stage to select the final sample units then it is called multistage sampling. When the final sample unit is selected in two stages then it is called two-stage sampling, if the final sample unit is selected in three stages then it is called three-stage sampling and so on. For example, if the researcher is going to conduct a survey of Nepalese people living in Nepal. If the researcher select district at the first stage, then VDC/Municipalities of the selected district in the second stage, wards of the sampled VDC/Municipalities in the third stage and household of the sampled wards in the fourth stage then it is called 4-stage or multistage sampling.

 

Sampling and non-sampling error:

 Sampling error:

Sampling error arises due to the imperfection of the sampling method of the survey. The difference between the sample statistics and population parameters due to the fluctuation of the sample is called the sampling error. This decreases as the sample size increases. This error is completely absent in the census method. The following are the reason for this type of error.

1. Faulty selection of a sample.
2. Substitution of sample units
3. Faulty demarcation of the sampling unit
4. Improper choice of statistics to estimate a population parameter.

Non-Sampling error:

This arises at the various stages of study in both census and sample survey. It is very difficult to control and identify the non-sampling type of error.  This error arises at the data collection, processing, analysis and publishing time of the report. The following are some reasons for this particular type of error.

1. Faulty planning and definition
2. Response error
3. Non-response error
4. Error in coverage
5. Compiling error
6. Publication error

You may also like: Non-probability Sampling Methods ||Estimation and Hypothesis