Sample Versus Population

The Basics of Populations and Samples

The art of sampling in communication research stems from the field of political science—more specifically, political polling. Starting in 1920, the Literary Digest sent postcards to voters in six states asking them questions such as who they planned to vote for in the upcoming election. In the 1936 election, the Digest sent an ambitions 10 million ballots to people listed in telephone directories and also on lists of automobile owners. Its poll predicted a landslide victory for Alf Landon over Franklin Roosevelt when in actuality Roosevelt won by the largest margin in history. To determine what went wrong, one only has to look to the sampling procedures. By polling only automobile owners and names listed in the phonebook, the Digest inadvertently targeted wealthier members of the population who were more likely to vote for the Republican candidate (Landon) rather than the Democratic candidate (Roosevelt). Given its lack of accuracy along with a poor response rate overall, it became clear that the Digest needed to strengthen how it sampled participants to enhance the accuracy of its results. Over the years, the method polling companies use to sample the population has improved dramatically in terms of sophistication. In presidential elections, polling companies like Gallup, Harris, and Pew are hired to examine political attitudes in an attempt to ultimately determine which candidate is in the lead, among other pieces of information. For Gallup, Harris, and Pew, the population would be all eligible voters in the United States. According to the U.S. Census Bureau, there are just more than 206 million eligible voters in the United States. It is not possible to survey every single person who is eligible to vote as it would be too costly and time consuming. Therefore, polling companies select a smaller sample that, if done effectively, is representative of the larger population of all voters. What follows is a discussion of defining the population.

Defining the Population

Once researchers have determined what they want to study they must first define their population and determine population boundaries. A population boundary separates who or what researchers are interested in studying from those they are not interested in investigating. For example, in a study investigating communication satisfaction in married couples following the birth of a child, researchers must make decisions about how they will narrow the larger population down to a more manageable group. Setting clear population boundaries will help define and further narrow the population. Using this same example, the population of all married couples can be narrowed to couples who have had a child. This is still a fairly large population, so the researchers may set additional population boundaries. For example, the researchers may decide to narrow their focus to same-sex married couples rather than opposite-sex married couples. The researchers could also set a boundary around the length of time that has passed since the child was born (e.g., 6 months to a year). The researchers could also focus on couples experiencing the birth of their first child. It is crucial for the researchers to set clear population boundaries and use specific language identifying the desired population to aid in fully identifying population members. In this case, using married couples is too vague as the researchers really want married couples who have experienced the birth of a child.

...