Random Sampling in Research Classes,

Random Sampling

In research classes, the importance of random sampling cannot be emphasized enough, especially when discussing empirical or quantitative methods of research. Random sampling is a systematic method of determining a quantitative study's unit of analysis (for example, a survey respondent) through the principle of 'having the equal chance to be selected' in the universe or target group of the study. It helps eliminate bias in the selection process of a study's respondent or unit of analysis, which could be the researcher himself/herself who could make the decision to include or exclude a potential respondent or unit of analysis. Random sampling also gives order to a research program in that it provides a technical guideline in creating order out of a seemingly disorganized 'universe.' For example, a survey enumerator would conduct a survey in an orderly manner through random sampling, rather than just interviewing individuals who are willing to be interviewed. Thus, through random sampling, researchers are ensured that units of analysis are objectively selected and can thus be claimed as representative of the universe, population, or group under study.

However, it could also be argued that random sampling could not be as important as it is supposed to be in quantitative studies. It has been a practice in market research, for example, to conduct purposive sampling to specifically target a group of interest. Purposive sampling specifically looks for an individual or unit of analysis given a set of criteria that must be followed; thus, if researchers are asked to search for this individual/unit of analysis given the set criteria, it would be very difficult to find this individual/unit of analysis using random sampling. Hence, purposive sampling is used. It is during these cases that a researcher could say that random sampling is not important, since in purposive sampling, the aim is to interview or survey the desired number of individuals/units of analysis sharing the same characteristics specified/identified by the researcher.

Another instance in which random sampling is not important is when a researcher is conducting a qualitative study. In qualitative studies, depth of information from selected units of analysis is given more importance rather than the number and representativeness of the target group or population. Thus, for a researcher conducting in-depth interviews of household mothers (for example), the characteristics/selection criteria set in identifying the target group (household mothers) are more important than the randomly selecting participants/members of the target group, which could result to household mothers who do not fit the criteria or characteristics set by the researcher. In effect, the type of information that would be generated from randomly sampled household mothers would not be as insightful as it would have been to specifically-selected household mothers fitting the researcher's selection criteria.