Pitfalls

There are many ways to select samples from the population. It is important to use a method that allows us to select a sample that is representative of the target population for reliable results to be obtained.
 

Obtaining a representative sample

Sometimes, we hear people mentioning that the sample selected for a survey is representative, as the demographic composition (e.g. age) in the sample closely matches that of the target population. However, this does not necessarily result in a representative sample. For example, suppose we would like to find out the monthly expenditure of people staying in a community. Selecting persons off a street to fulfill a specified number of young and old, male and female respondents based on the target population composition might not give us a representative sample. Certain groups of people may be left out; e.g. high-income earners who normally do not pass by that street. This could bias the survey results.

Estimates (e.g. mean age) from any two representative samples of the same size, selected from the same population using the same sampling method, could differ due to sampling and non-sampling errors. Sampling errors arise when we make conclusions about a population based on the findings of a sample of this population. On the other hand, non-sampling errors could arise when, e.g. some respondents provided inaccurate information because they interpreted the question wrongly. Details can be found at the page on “How Reliable is a Sample?”.