In Statistical inquiry, the investigator has the choice of collecting the relevant data from either all the units in the affected group or from a part thereof.
Sampling can be defined as the selection of items from a given population in such a way as to secure data on the basis of a reasonable conclusion can be drawn regarding the entire population.
Advantages of Sampling
- Reduced Cost: If data were collected from a fraction of the population, the expenditure would be smaller than when a census is attempted.
- Follow-up: follow-ups of non-responses are easier and cheaper in most cases.
- Errors: Sampling makes for easy assessment of errors in any investigation.
- Further statistical analysis: Whereas Census only allows for the computation of the population parameters, sampling gives room for further analysis of the results from the sample
Non-probability Sampling Techniques
- Accessibility (Convenience) Sampling: This is a non-probability sampling method where the more convenient or accessible elementary units are chosen from the population for observation.
- Purposeful (Judgemental) Sampling: This is obtained according to the discretion of someone who is familiar with the relevant characteristics of the population.
- Quota Sampling: In this case, the interviewer is free to choose his samples as he pleases provided the quota requirement is met. By this method, Interviewers working in particular areas are instructed to interview.
Disadvantages of Sampling
- Most correspondents either decline to give answers to questions, and those who consent to give information may come up with hurried answers only.
- It gives rise to a great deal of bias.
Statistical sampling is based on the assumption that the selection of the sampling units has been done in a random manner.
- Sampling with Replacements: This refers to a situation in which the element selected into the sample is returned into the population after every necessary examination before the next selection is effected. so that the population size remains the same at every drawing.
- Sampling without Replacements: In this approach, every element selected for examination from the population is removed from the population after the examination.