The usefulness of any research is dependent upon how well the group studied represents the group about which decisions are to be made or conclusions drawn. That is, it depends upon how well the sample reflects relevant characteristics of the population. When it is possible to study every member of that group there is no problem, for on these occasions we can easily calculate the exact attribute (parameter) of interest for our population.
For example, if we were interested in determining the average number of gallons of gasoline sold to customers at our service station yesterday, we could add up the total number of gallons sold yesterday and divide that by the total number of gas purchases. The result of this census would be the exact average number of gallons of gasoline sold per customer (yesterday).
But when it is not possible to study each and every member of the target population, researchers and analysts must estimate characteristics of the group based upon the study of a relatively small number of elements. This smaller group is referred to as a sample, and the process of selecting the individuals who will be included in the sample is called sampling.
If we wanted to determine the average number of gallons of gasoline sold to customers over the last year, we would be unlikely (or unwilling) to perform a census of all gasoline sales over such a long period. It would be too time consuming and/or costly to find and drag out the records for each and every day of an entire year and make the calculations.
In this situation we would be more likely to take a sample from among all of the gasoline purchases made during the prior year. One simple method of taking this sample would be to take a random group of, say, 50 days from the past year. The number of gallons sold for each sale over each of the 50 days could then be added together and divided by the total number of individual sales over those 50 days.
The resulting statistic is called the sample mean. The sample mean (along with other information) may then be used to estimate the average number of gallons per purchase for all gasoline purchases during that year. This process of estimating a population parameter from a sample statistic is called inferential statistics.
This paper deals with issues surrounding the process of determining the appropriate sample for a research study and the factors you will need to check before you are able to conclude that the sample used is appropriate for both the analysis performed and the management decision you need to make.
Boyd, P. (2002). Sampling concepts. Unpublished manuscript, The Alan Shawn Feinstein Graduate School, Johnson & Wales University, Providence, Rhode Island.
Boyd,, Paul Ph.D., "Sampling Concepts" (2002). MBA Faculty Conference Papers & Journal Articles. Paper 2.
Accounting Commons, Advertising and Promotion Management Commons, Agribusiness Commons, Business Administration, Management, and Operations Commons, Business and Corporate Communications Commons, Business Law, Public Responsibility, and Ethics Commons, Corporate Finance Commons, E-Commerce Commons, Entrepreneurial and Small Business Operations Commons, Finance and Financial Management Commons, Food and Beverage Management Commons, Gaming and Casino Operations Management Commons, Human Resources Management Commons, Insurance Commons, International Business Commons, Labor Relations Commons, Management Information Systems Commons, Management Sciences and Quantitative Methods Commons, Marketing Commons, Organizational Behavior and Theory Commons, Other Business Commons, Portfolio and Security Analysis Commons, Real Estate Commons, Recreation Business Commons, Sales and Merchandising Commons, Strategic Management Policy Commons, Taxation Commons, Technology and Innovation Commons, Tourism and Travel Commons