Random Sampling Why Random Sampling

Random Sampling

Why random sampling is not important

Random sampling is a statistical and research concept wherein a sample of a population is determined through objective and systematic methods -- that is, through probabilistic methods. The probabilistic or random sampling method is considered the most preferred and superior form of sampling, particularly in quantitative studies. It ensures that no bias or objective judgments were made during the selection of the sample and that each unit from a target population was given an equal chance of being 'randomly' selected as part of the sample for the study. It also reduces the risk of error during sample selection, which helps increase the confidence level of the study results.

Despite these advantages of random sampling to research, there are also instances when probabilistic sampling method is not important in a research study. This is especially true in studies with qualitative design. Under the qualitative design, the researcher puts more importance to his/her subjective judgment as an expert on the topic under study, therefore objective and systematic method of selecting a sample is not critical in this case. A qualitative researcher would use non-probabilistic methods of sampling, which includes selecting units for the sample that would provide more depth and dimension to the concept or phenomenon being studied. With quantitative studies and random sampling, representativeness and generalized results about the general population or group under study are the primary concern. Qualitative studies and non-probabilistic sampling, meanwhile, puts more emphasis on the richness and validity of information that are culled out from subjectively selected units of a highly specific / uniquely identified sample.

However, there are also cases when even quantitative studies have to resort to non-probabilistic means of determining units to be included in the sample. There are sampling methods that are still under the quantitative paradigm but utilize non-random sampling to collect data from samples that are either too expensive or logistically difficult to select if random sampling will be used instead. These sampling methods can be convenience, purposive, quota, and snowball sampling (Babbie, 2003:211).

Convenience sampling is a non-random sampling method used when the goal is just to achieve the target sample size of a specific group for the study. No other criteria are set except for the kind of group being targeted and sample size that needs to be achieved. Purposive sampling, meanwhile, specifies characteristics of units that will be included in the sample. These units could be extreme characteristics of each other, or maybe only units from the extreme end of a characteristic spectrum are needed. The objective of this sampling method is to include units of a specific type or a combination of different types. Quota sampling is used when the researcher's objective is to achieve a specific proportion of the target population. Snowball sampling, lastly, is used when the units of a sample are difficult to reach, such as commercial sex workers, HIV-infected individuals, substance abuse victims, among many others.