A researcher is conducting research on SMEs where he selected the industrial areas based on judgmental sampling and there after he approached the government training centre of each industrial area for registered SMEs data, collected the SMEs record and accordingly conducted the primary data survey among the SMEs.
Kindly let me know the type of sampling technique used in the above process.
future subjects (sampling units) were not selected from from among the acquaintances. In addition, researcher collected the record of SMEs coming for training and theirafter he conducted the survey.
Kindly help me out in identifying the right sampling technique.
It is non-probability sampling and can be treated as Purposive sampling.
@Professor Sushil,
Sir, In case if we do not have complete database for sampled units. How to conduct the primary data survey.
In your paper, you will need to inform the readers how many industrial areas are there in total and how many you selected using judgemental sampling. Secondly, you focused on registered SMEs and worldwide it is known that majority of SMEs do not register with Government. In short, your research findings will be restricted to those cases selected and may not be generalized. You can go ahead and collect primary data from those on the register. Maybe your findings will encourage more SMEs to register in the future.
If the "industrial areas" had been chosen at random, that would have been the beginning of cluster sampling. If the selected clusters (industrial areas) had then been further subject to a randomized design, then it would have been a two-stage cluster sample. But as Albert Otete stated, your results "may not be generalized." Such inference is only possible through randomized designs, or perhaps when regressor data are available on the entire population where prediction (regression models) may be appropriate. A model should only be used for a set of data for which it is appropriate.
The following may be of interest to many readers:
Waksberg Award article:
Brewer, K.R.W. (2014), “Three controversies in the history of survey sampling,” Survey Methodology,
(December 2013/January 2014), Vol 39, No 2, pp. 249-262. Statistics Canada, Catalogue No. 12-001-X.
http://www.statcan.gc.ca/pub/12-001-x/2013002/article/11883-eng.htm
(The example found here for the classical ratio estimator is not the only model Ken Brewer used, but is used for illustration here.)
Since you do not seem to have a list of population members, then area sampling, which is cluster sampling where you use areas, such as city blocks, or say your "industrial areas," would seem your best option in the future. You would need to pick the industrial areas at random, and then either be able to census the SMEs in each selected area, or do a further randomized design for SMEs within each area. The random sampling is not restricted to simple random sampling, but could get rather complicated otherwise.
There are many good sampling books. Here is a classic:
Cochran, W.G(1977), Sampling Techniques, 3rd ed., John Wiley & Sons.
- Badges
- Science topic
- Similar topics
- Organizational Psychology
- Training