A sampling frame is a list of all eligible members of a population from which samples are drawn. It can be thought of as the pool from which samples are obtained. It is a statistical framework used in surveys, social research, marketing research, and different types of studies. This frame is necessary to arrive at an unbiased, accurate conclusion or finding because it defines the population being studied completely. It's not normally possible or even practical to make direct observations of every element in the population of interest, and a frame restricts the population being studied to a manageable figure, ultimately helping the researchers to draw conclusions about the entire population.
For instance, a study aimed at figuring out the time teenagers spend online can't really include every teenager in the world. Certain parameters are introduced to make the population of interest smaller. A sampling frame in this case might specify that the teenager live in and around New York, be between the ages of 13 and 15, has access to a computer at home, and attends a public school. A study conducted this way may hope to bring out insights that apply to teenagers in general in this segment. Establishing a clear frame is critical to the success of any survey or study as a faulty frame leads to inconsistent or inaccurate findings or results.
Though the frame narrows down the pool from which the sample is drawn, it differs from the population of interest to some extent. For instance, using the above example, the sampling frame doesn't include teenagers who access the web from their mobile phone, who aren't at home at the time of the call, or who simply aren't interested in participating in surveys. Even getting into the sampling frame doesn't ensure that the person becomes a part of the final sample group. Samples may be drawn randomly from the frame where every person has a chance of being included or in a more systematic fashion, say, when every tenth person in the list is selected.
There are numerous problems, which those drawing up a sampling frame experience, that may skew the results. Missing members are a very common problem where those who need to be within the frame have been left out by accident. Duplicate members are also a big issue, where a member is listed more than once. Sometimes foreign entries — people who don't represent the population of interest — can be found within the frame. Other times, instead of individuals being listed, the frame may contain groups. When mistakes exist within the sampling frame, the final sample drawn is faulty, either as a sample that is unrepresentative of the group being studied or containing significant bias.
