Following the American Psychological Association’s Guidelines
When a researcher wants to observe, look into or test a theory or hypothesis, he will consider the problem by selecting a section of the study’s population through a method called random sampling. In random sampling, all the subjects of the population identified in the study take on the same chances of getting selected for the sample. This means that every member of the sample essentially retains the characteristics, or unbiased characteristics, of the population. With random sampling, the conclusions reached after hypothesis testing that apply to the sample selection will apply for the entire population as well. This is because the sample selection essentially represents the characteristics of the population from where it is obtained, as each member of the sample was drawn from the population data in an unbiased way. When bias is avoided in the sample selection, the results generated by a particular study are viewed to be more conclusive and error is minimized.
When random sampling is used in research, each member of the target population is identified and assigned a random number so that each will have equal chances of getting selected for the sample. This can prove to be a costly, very lengthy and time-consuming process. In some cases, it may not be possible to randomly assign a number when the target population is large. This is very true in simple random sampling. A researcher can opt to use alternative random sampling methods such as stratified sampling or systematic sampling or even cluster and stage sampling. These methods will ensure a more representative sample data without sifting through all population data in detail. When the main concern is to remove bias in the sample data, a good way to do so may be to increase the sample size. Again, there are several random sampling methods to choose from. Any of these alternative random sampling techniques should yield results representative of the population. The problem inherent in simple random sampling method should be dealt with by restructuring the research methodology to allow a more economical and less time-consuming way to select a sample for the study.
References
American Association of Public Opinion Research. (2013). What is a Random Sample. Retrieved from http://www.aapor.org/What_is_a_Random_Sample_1.htm
Holah.co.uk. (n.d.). Sampling. Retrieved from http://www.holah.co.uk/page-detail.php?slug=sampling.
Research Methods Knowledge Base. (2006). Probability Sampling. Retrieved from http://www.socialresearchmethods.net/kb/sampprob.php.
