Population Analysis Essay

754 Words 4 Pages
As the owner of a popular clothing establishment, it came to my attention during the weekly staff meeting that customer satisfaction was a concern the employees were voicing. This discussion led to the potential of performing a survey throughout our ten locations, which serve over two-hundred customer per day. The quandary is how we build this sample.
When gathering satisfaction data from a population, such as the customers of a popular clothing store, surveys are very beneficial. However, there are two different survey methodologies used, which are population studies and sampling. Both methods can provide meaningful as well as accurate data, but noticeable limitations can be encountered as well as advantages gained from these methods. Thus,
…show more content…
The first step in defining a population is to decide whether it is a population of individuals, households, or whatever. Once you establish the unit of analysis, then decisions are made as to what units should be included or excluded in the survey. This is generally a consideration of demographic criteria such as location or area, the age of individuals, as well as, a plethora of other demographic variables that include sex, race, marital status, and education that are useful in defining a population. In addition, if the unit of potential analysis is a household, then one define the variables li such as adults, teens, children, or …show more content…
Not only is doing a sample of a population more convenient, but it may also ensure better cooperation from those individuals included in the survey, thus, the goal becomes to get a representative sample, which is a sample that provides a lesser picture of all aspects in the bigger picture, while, being proportionate. Thus, a sample is not just a part of the population studied, but an accurate miniature of the population under consideration. If a typical selection is not obtained, then the sample is insufficient to represent the entire population, thus might not be sufficient to warrant statistical analysis. Hence, if a valid random sample is drawn, as well as, the required number of responses gathered, the results can represent, with a ninety to ninety-nine percent confidence level, the total population one attempts to survey (MIT, 2011). A specific sample target number has to be determined at the onset of the survey process to ensure that the final results meet the required confidence level

Related Documents