Nielsen determines its top 100 ratings list through statistical sampling, a process that surveys a small, representative sample of the overall population to estimate the number of actual viewers for each program. The process helps determine advertising rates and which TV shows get cancelled.
Nielsen?s main measure of viewing comes from about 5,000 households selected to be the representative sample for national estimates. Because this sample is much smaller than the actual TV audience, Nielsen tries to make this sample as representative as possible of the overall population of U.S. viewers. The households selected represent a cross-section of American demographics, with significant variations in age, income, ethnicity and household size. Taking information from these sample households, Nielsen estimates actual viewership numbers based on the total population of U.S. television viewers.
To actually collect data, Nielsen provides sample households with television meters known as ?black boxes.? These devices are simply Internet-connected computers that log viewing data from home TV sets. Sample household members also have individual devices that they switch on when they are watching television, so Nielsen knows who in the household is watching what.
Television networks assure accuracy in these ratings by performing audits, quality checks and their own viewing surveys. Nielsen also compares information against larger surveys of the general population.