For more than 50 years, the National Association of REALTORS® has been dedicated to fundamental and meaningful research. NAR uses the following methods and processes to conduct data collection and analysis.
Sampling is a method used in statistics to efficiently analyze a general population. While it is difficult to survey the entire population, researchers use sample groups to conduct their inquiries. For that reason, sampling frequently requires random selection in order to create a more representative sample. For instance, to find out how members feel about a new issue or proposal, NAR would send the survey to either 50,000 random members or a target population of 50,000 residential or commercial members.
In some cases, NAR researchers may need to oversample a group or receive higher responses from those in a particular state or who are passionate about a particular issue. Accordingly, a sample may be weighted to more accurately represent the population as a whole. For example, for the Member Profile, NAR oversamples states and locals who are interested in a proprietary report. If more responses are received from a particular state than from other states, that state is down-weighted to prevent it from being overrepresented.
A margin of error is used to ensure confidence in sampling. NAR researchers strive to have margins of error under +/- 2.0 percent. For smaller populations, they focus on the margin of error to ensure reliability and accurate consideration as quantitative data.
Both REALTOR® members and American households are asked to participate in surveys frequently. For surveys of members, the typical sample size is at least 2,000 responses. NAR conducts over 150 surveys annually. For example, one of the largest surveys NAR conducts every year is the Profile of Home Buyers and Sellers survey, which is mailed to recent buyers who purchased a home in the last year.