Curious to know why some shows stay on television forever, while some of your favorites are gone too soon? This is largely due to the show's ratings--a random sampling of thousands of people used to estimate how many and what types of people are viewing each program at any given time. These are compiled by Nielsen Media Research, which uses a variety of means to collect the necessary information.
Nielsen has been doing customer and advertising research since the 1920s and has been gauging national TV viewership since 1950. Today, it remains the main source of ratings data for every broadcast and cable network.
Nielsen conducts random sampling of people across the country to select a broad range of subjects from different races, ages and locations, as well as different economic and demographic groups (single, married, with children) to provide as true an estimate as possible.
The primary way Nielsen measures viewership is through meters in approximately 25,000 households in the United States. These record on a minute-by-minute basis what is being viewed in that home and by which members of the household. That information is relayed back to Nielsen each night for the overnight ratings.
Nielsen also engages additional people (around 1.6 million) to keep notebook diaries at four different "sweeps" months during the year. The diaries are for a period of one week and are mailed back to Nielsen upon completion to be added into the cumulative ratings.
Nielsen has also begun collecting information directly from certain cable providers, including Charter. The data is collected and relayed through the customers' set-top receiver boxes and digital video recorders.
The combination of information sources is analyzed and distributed to each network, giving them a scientific viewership estimate with which to determine advertising rates and judge the value and fate of their programs.