AAAA (American Association of Advertising Agencies): trade association of U.S. advertising agencies.
Active Audience: term given to viewers who are highly selective about the programming they choose to watch. Active audiences are sometimes defined as those who turn a set on only to watch favored programs, and turn the set off when those programs are unavailable. Activity can also mean being goal-directed in media selections, or cognitively engaged with the media. See LOP, passive audience.
Adjacency: advertising opportunity immediately before or after a specific program.
ADI (Area of Dominant Influence): term once used by Arbitron to describe a specific market area. Every county in the United States was assigned to one, and only one, ADI. See DMA.
Advertising Agency: company that prepares and places advertising for its clients. Agencies typically have media departments that specialize in planning, buying, and evaluating advertising time.
Adware: program secretly placed on a user’s computer that redirects the browser to selected sites and/or launches pop-up ads.
Affiliate: broadcast station that has a contractual agreement to air network programming.
Algorithm: computational procedure in which data are reduced, usually through a number of steps. Algorithms are used to produce audience metrics and recommendations to media users.
AMOL (Automated Measurement of Lineups): system that electronically determines the broadcast network programs actually aired in a local market.
ANA (Association of National Advertisers): trade organization of major national advertisers responsible for creating the first broadcast ratings service. See CAB.
A/P Meter (Active/Passive Meter): type of peoplemeter developed by Nielsen that has the ability to either detect a code embedded in a television program or actively identify its audio signature, thus identifying the programming on the set.
App: short for “application,” it is a piece of software designed to help users perform a specific task (e.g., play a game, fetch a particular type of information, etc.). Apps are popular on mobile devices like smartphones and tablets.
AQH (Average Quarter Hour): standard unit of time for reporting average audience estimates (e.g., AQH rating, AQH share) within specified dayparts.
ARB (Audience Research Bureau): ratings company established in 1949 that was the predecessor of the Arbitron Company.
Arbitron: major supplier of local market and national network radio ratings.
Area Probability Sample: type of random sample in which geographic areas are considered for selection in some stage of the sampling process. See probability sample, cluster sample.
ARF (Advertising Research Foundation): trade organization of advertising and marketing research professionals in the United States and Canada advancing the practice and validity of advertising research.
Arianna: software program that handles television ratings; joint venture between Nielsen and AGB.
Ascription: procedure for resolving confused, inaccurate, or missing data entries.
Audience Appreciation (AA) data: supplemental data collected by some TAM systems that asks viewer to rate how much they like or appreciate a program. AA data are often collected through meters, though other survey techniques can be used.
Audience deficiency (AD): failure to deliver the numbers and kinds of audiences agreed to in a contract between time sellers and buyers. Sellers will often remedy audience deficiencies by running extra commercials, called “make-goods.”
Audience duplication: cumulative measure of the audience that describes the extent to which audience members for one program, channel, or website are also in the audience of another program, channel, or website. Audience duplication data are a building block for many forms of cumulative audience analysis. See audience flow, channel loyalty, inheritance effect, repeat viewing, recycling.
Audience flow: extent to which audiences persist from one program or time period to the next. See audience duplication, inheritance effects.
Audience fragmentation: phenomenon in which the total audience for a medium is widely distributed across a large number of outlets. Cable is said to fragment the television audience, resulting in a decreased average audience share for each channel. Audience fragmentation contributes to problems of sampling error.
Audience polarization: phenomenon associated with audience fragmentation, in which the audiences for outlets or specific types of content use them more intensively than an average audience member. See channel loyalty, channel repertoire.
Audience turnover: phenomenon of audience behavior usually expressed as the ratio of a station’s cumulative audience to its average quarter-hour audience.
Audimeter: Nielsen’s name for several generations of its metering device used to record set tuning. See SIA.
Availabilities: advertising time slots that are unsold, and therefore available for sale. Sometimes called avails.
Available audience: total number of people who are, realistically, in a position to use a medium at any point in time. It is often operationally defined as those actually using the medium (i.e., PUT or PUR levels).
Average: measure of central tendency that expresses what is typical about a particular variable. An arithmetic average is usually called a mean. See mean, median.
Average audience rating: rating of a station or program during an average time interval over a specified period of time. Metered data, for example, allow reports of audience size during an average minute of a television program.
Average time per page: measure of the average time spent with a Web page across however many pages are examined in a single visit.
Away-from-home listening: estimates of radio listening that occurs outside the home. Such listening usually takes place in a car or place of work. Also called out-of-home.
Banner advertising: form of display advertising on the Internet, in which a box containing the advertiser’s message appears on a portion of the page being viewed. Banner advertising often allows users to be linked to the advertiser’s Web site.
BARB (Broadcasters’ Audience Research Board): A joint industry committee that oversees the production of TAM data in the United Kingdom. See JIC.
Barter: type of program syndication in which the cash expenditure for programming is reduced, sometimes to zero, because it contains national or regional advertising that is sold by the syndicator.
Basic cable: programming services provided by a cable system for the lowest of its monthly charges. These services typically include local television signals, advertiser-supported cable networks, and local access.
BBM Canada: nonprofit entity established by Canadian broadcasters and advertisers to conduct audience measurement.
Birch: research company that once provided syndicated radio rating reports in competition with Arbitron.
Block programming: practice of scheduling similar programs in sequence in order to promote audience flow. See inheritance effect.
Bounce: tendency of a station or network’s ratings to fluctuate over time due to sampling error, rather than real changes in audience behavior. Bounce is most noticeable for outlets with low ratings.
Bounce rate: is the percent of website visitors who leave without taking any further action on the site (e.g., clicking on another page).
Broadband: term describing the channel capacity of a distribution system. A common label for multichannel cable service, it is also applied to digital networks capable of delivering full motion video. See cable system.
Browser: computer program that allows users to gain access to pages on the World Wide Web. Browser can be set to accept or reject cookies, which are important in measuring Web use. See cookies.
Buffer sample: supplemental sample used by a rating company in the event that the originally designated sample is insufficient due to unexpectedly low cooperation rates.
C3: See commercial rating.
Cable Advertising Bureau (CAB): U.S. trade organization formed to promote advertising on cable television.
Cable penetration: extent to which households in a given market subscribe to cable service. Typically expressed as the percent of all television households that subscribe to basic cable.
Cable system: video distribution system that uses coaxial cable and optical fiber to deliver multichannel service to households within a geographically defined franchise area.
Call-back: practice of attempting to interview someone in a survey sample who was not contacted or interviewed on an earlier try. The number of callback attempts is an important determinant of response rates and nonresponse error. See nonresponse error.
Cash-plus-barter: type of barter syndication in which the station pays the syndicator cash, even though the program contains some advertising. See barter.
CATV (Community Antenna Television): acronym for cable television, used in many early FCC proceedings.
Census: study in which every member of a population is interviewed or measured. Every 10 years, the federal government conducts a census of the U.S. population. Server-centric measurement sometimes claims to provide a census of media use.
Channel loyalty: common feature of television audience behavior in which the audience for one program tends to be disproportionately represented in the audience for other programs on the same channel. See audience duplication, inheritance effects.
Channel repertoire: set of channels from which a viewer chooses—typically much fewer than the total number of channels available.
Cinema Advertising Council (CAC): U.S. trade organization to promote the sale of commercial time in theaters.
Circulation: total number of unduplicated audience members exposed to a media vehicle (e.g., newspaper, station) over some specified period of time. See cume, reach.
Clearance: (1) the assurance given by a station that it will air a program feed by its affiliated network; (2) the sale of syndicated programs to individual markets.
Click: when the user of a Web page interacts with (i.e., clicks on) a message.
Click fraud: any method of artificially inflating the number of clicks on a Web site.
Click rate: percentage of advertising responses as a function of the number of clicks.
Clickstream: record of all http requests made from a browser.
Click-through: or click-through-rate (CTR) is a measure of Web ad effectiveness that tracks how many Web users actually click on the ad they see.
Cluster sample: type of probability sample in which aggregations of sampling units, called clusters, are sampled at some stage in the process. See probability sample.
CMSA (Consolidated Metropolitan Statistical Area): type of metropolitan area, designated by the U.S. Office of Management and Budget, often used by ratings companies to define a media market’s metropolitan area.
Codes: (in survey research) the numbers or letters used to represent responses in a survey instrument like a diary. Coding the responses allows computers to manipulate the data.
Cohort: type of longitudinal survey design in which several independent samples are drawn from a population whose membership does not change over time. See longitudinal.
Coincidental: type of phone survey in which interviewers ask respondents what they are watching or listening to at the time of the call. Coincidentals, based on probability samples, often set the standard against which other ratings methods are judged.
Collaborative Filtering: sophisticated way to make recommendations that uses computer algorithms to compare a person’s past actions to those of users with a similar profile.
COLRAM (Committee on Local Radio Audience Measurement): committee of the NAB concerned with a range of local radio measurement issues.
COLTAM (Committee on Local Television Audience Measurement): committee of the NAB concerned with a range of local television measurement issues.
COLTRAM (Committee on Local Television and Radio Audience Measurement): committee of the NAB which, in 1985, was divided into COLRAM and COLTAM.
Commercial rating: measure of the number of people or households who have been exposed to commercial messages. In the United States commercial ratings, called C3, are an average across all commercials in a program plus three days of replay.
Confidence interval: in probability sampling, it is the range of values around an estimated population value (e.g., a rating) with a given probability (i.e., confidence level) of encompassing the true population value.
Confidence level: in probability sampling, it is a statement of the likelihood that a range of values (i.e., confidence interval) will include the true population value.
Convenience sample: nonprobability sample, sometimes called an accidental sample, that is used because respondents are readily available or convenient.
Conversion rate: number of desired actions (e.g., placing an order on a website) divided by the number of visits or unique visitors.
Cookies: small text file sent by a server to a user’s browser that serves to identify users and provide a record of their actions. Cookie counts help measure unique visitors, but must typically be adjusted with algorithms to avoid over-or under-counting users. They also provide a basis for serving targeted advertising. Cooperative Analysis of Broadcasting (CAB): first ratings company in the United States. Formed in 1930 by Archibald Crossley, it ended operations in 1946.
Correlation: statistic that measures the strength and direction of the relationship between two variables. It may range in value from +1.0 to −1.0, with 0 indicating no relationship.
Counterprogramming: programming strategy in which a station or network schedules material appealing to an audience other than the competition. Independents often counter-program local news with entertainment.
Coverage: potential audience for a given station or network, defined by the size of the population that is reached, or covered, by the signal.
CPA (cost per action): measure the cost of an Internet ad based on how many users take some action in response to an ad divided by the cost of the ad.
CPI (cost per impression): measure of the cost of an ad based on the number of impressions divided by the cost of the ad.
CPM (cost per thousand): measure of how much it costs to buy 1,000 audience members delivered by an ad. CPMs are commonly used to compare the cost efficiency of different advertising vehicles.
CPP (cost per point): measure of how much it costs to buy the audience represented by one rating point. The size of that audience, and therefore its cost, varies with the size of the market population on which the rating is based.
Cross-sectional: type of survey design in which one sample is drawn from the population at a single point in time. See longitudinal.
Cross-tabs: technique of data analysis in which the responses to one item are paired with those of another item. Cross-tabs are useful in determining the audience duplication between two programs. See audience duplication.
Cume: short for cumulative audience, it is the size of the total unduplicated audience for a station over some specified period of time. When the cume is expressed as percent of the market population it is referred to as cume rating. See circulation, reach.
Cume duplication: percentage of a station’s cume audience that also listened to another station, within some specified period of time. See exclusive cume.
Daypart: specified period of time, usually defined by certain hours of the day and days of the week (e.g., weekdays vs. weekends), used to estimate audience size for the purpose of buying and selling advertising time. Dayparts can also be defined by program content (e.g., news, sports).
Data fusion: method of combining data from two of more separate samples by pairing individual respondents based of similarities in their sociodemographic and behavioral profiles. Data fusion is often used as an alternative way of studying cross-platform media use when no single source dataset is available.
Demographics: category of variables often used to describe the composition of audiences. Common demographics include age, gender, education, occupation, and income.
Diary: paper booklet, distributed by ratings companies, in which audience members are asked to record their television or radio use, usually for 1 week. The diary can be for an entire household (television) or for an individual (radio).
Direct Broadcast Satellite (DBS): mode of television distribution that uses signals transmitted via satellite to provide programming directly to subscriber households.
Display advertising: category of advertising. In newspapers it includes larger advertisements run by retailers and is often contrasted with classified ads. On the Internet, it includes banner ads and is often contrasted with paid search advertising.
DMA (Designated Market Area): term used by Nielsen to describe specific market areas in the United States. Every county belongs to one, and only one, DMA.
Domain consolidation level: consolidation of multiple domain names and/or URLs associated with a main site.
Domain name level: consolidation of multiple URLs associated with the same domain name.
DST (differential survey treatment): special procedures used by a ratings company to improve response from segments of the population known to have unusually low response rates. These may include additional interviews and incentives to cooperate.
DVR (digital video recorder): electronic device that records video programming on a hard drive. Sometimes called a personal video recorder (PVR), it empowers viewers with the ability to easily record, fast forward, replay programming, and skip commercials. TiVo is the best known DVR system.
EACA (European Association of Communications Agencies): European trade association of advertising agencies and related media specialists.
Early fringe: in television, a daypart in late afternoon immediately prior to the airing of local news programs.
EBU (European Broadcasting Union): An international trade association of radio and television broadcasters in over 50 countries across Europe, North Africa, the Middle East and other regions.
Editing: procedures used by a ratings company to check the accuracy and completeness of the data it collects. Editing may include techniques for clarifying or eliminating questionable data. See ascription.
Effective exposure: concept in media planning stipulating that a certain amount of exposure to an advertising message is necessary before it is effective. Often used interchangeably with the term effective frequency. See frequency.
Engagement: broad, ill-defined, term encompassing measures designed to assess user’s liking of or involvement with various media products and services. Sometimes these are derived from more conventional measures of exposure (e.g., repeat viewing or time spent metrics). Sometimes they are derived from actions taken on social media sites (e.g., comments, sharing, etc.). Sometimes engagement is measured directly with dedicated surveys.
Enumeration survey: large-scale survey of the population designed to provide population estimates that can be used to adjust or weight data provided by ratings panels.
ESF (expanded sample frame): procedure used by Arbitron to include in its sample frame households whose phone numbers are unlisted. See sample frame.
ESOMAR (European Society of Opinion and Marketing Research): global association of marketing professionals and opinion pollsters.
ESS (effective sample size): size of a simple random sample needed to produce the same result as the sample actually used by the rating company. ESS is a convenience used for calculating confidence intervals. Also called effective sample base, or ESB.
Ethnography: term that describes any one of several qualitative research techniques. Audience ethnographies include in-depth interviews or group discussions, studying what people post on the Internet, and a variety of observer and participant observer techniques.
Exclusive cume audience: total size of the unduplicated audience that listens exclusively to one station within some specified period of time.
Exit rate: measure of how many visitors left a website from a particular page.
Fault rates: measure of the extent to which installed meters are not providing useable data. Faulting can result for hardware failures or respondents failing to enter data.
FCC (Federal Communications Commission): U.S. independent regulatory agency, created in 1934, that has primary responsibility for the oversight of broadcasting and cable.
Format: style of programming offered by a radio station. Common formats include MOR (middle of the road), news/talk, and adult contemporary.
Frequency: in advertising, the average number of times that an individual is exposed to a particular advertising message.
Frequency distribution: way of representing the number of times different values of a variable occur within a sample or population.
Fringe: in television, dayparts just before prime time (early fringe) and after the late news (late fringe).
Fusion: See data fusion.
Geodemographics: type of variable that categorizes audiences by combining geographic and demographic factors. For example, organizing audiences by zip codes with similar population age and income.
Grazing: term describing the tendency of viewers to frequently change channels, a behavior that is presumably facilitated by remote control.
Gross impressions: total number of times an advertising schedule is seen over a period of time. The number of gross impressions may exceed the size of the population, since audience members may be duplicated. See GRP.
Group quarters: dormitories, barracks, nursing homes, prisons, and other living arrangements that do not qualify as households, and are, therefore, not measured by ratings companies.
GRP (gross rating point): gross impressions of an advertising schedule expressed as a percentage of the population. GRPs are commonly used to describe the overall size or media weight of an advertising campaign. GRPs = reach × frequency.
Hammocking: television programming strategy in which an unproven or weak show is scheduled between two popular programs in hopes that viewers will stay tuned, thereby enhancing the rating of the middle program. See audience flow, inheritance effect.
Headend: part of a cable system that receives television signals from outside sources (e.g., off-the-air, satellite) and sends them through the wired distribution system. See cable system.
Head of household: common, if somewhat dated, term used in TAM that denotes the adult with main responsibility for the household. By convention, in married households, it is the husband. There is only one head per household.
Hit: once common, gross measure of website popularity. A hit occurs anytime a user requests a file from a Web server. Since a single Web page might contain many different files, counting hits can greatly overstate page views. The number of visits, visitors, or page views are now more widely accepted as measures of popularity.
Home county: county in which a station’s city of license is located.
Home market: market area in which a station is located.
Home station: any station licensed in a city within a given market area.
Household: identifiable housing unit, such as an apartment or house, occupied by one or more persons. See group quarters.
HPDV (households per diary value): number of households in the population represented by a single diary kept by a sample household. Used to make audience projections. See projected audience.
HUT (households using television): term describing the total size of the audience, in households, at any point in time. Expressed either as the projected audience size, or as a percent of the total number of households.
Hypoing: any one of several illegal practices in which a station, or its agent, engages in an attempt to artificially inflate the station’s rating during a measurement period. Also called hyping.
IAB (Interactive Advertising Bureau): U.S. trade association promoting the Internet and other interactive technologies as advertising media.
Impression: basic gross measure of audience. It is the counts the number of times a program or ad is seen by audience members. Online, it is the number of times an Internet ad is successfully served to a user’s browser.
Independent: commercial television station that does not maintain an affiliation with a broadcast network.
Inertia: description of audience behavior that implies viewers are unlikely to change channels unless provoked by very unappealing programming.
Inheritance effect: common phenomenon of television audience behavior, in which the audience for one program is disproportionately represented in the audience of the following program. Sometimes called lead-in effects, audience inheritance can be thought of as a special case of channel loyalty. See audience duplication, audience flow, channel loyalty.
In-tab: term describing the sample of households or persons actually used in tabulating or processing results.
Internet: network of computer networks around the world that makes possible services like e-mail and the World Wide Web.
Interview: method of collecting data through oral questioning of a respondent, either in person, or over the phone.
Interviewer bias: problem of introducing systematic error or distortions in data collected in an interview, attributable to the appearance, manner, or reactions of the interviewer. See response error.
IP address: bit of computer code specified under an “Internet protocol” identifying your computer to servers on the Internet. IP addresses can be static, meaning they never change, or dynamic, meaning they are assigned each time you log on.
ISP (Internet Service Provider): company providing access to the Internet. These can include dial-up services, phone companies providing DSL and mobile devices, and cable systems offering high-speed modems.
JIC (Joint Industry Committee): common way to arrange for audience measurement services outside the United States. Affected industries (e.g., the media and advertisers), create a committee that specifies what services are needed, bids and awards a multiyear contract.
Keyword search: kind of advertising on the Web in which advertisers bid to have their site appear in response to a user’s search request. Also called pay-for-placement or paid search advertising.
Late fringe: in television, a daypart just after the late local news (11 P.M. EST).
Lead-in: program that immediately precedes another on the same channel. The size and composition of a lead-in audience is an important determinant of a program’s rating. See inheritance effect.
Lead-in effect: See inheritance effect.
Linear media: media systems that deliver content on a fixed schedule. Broadcast radio, television and cable are linear delivery systems. These are in contrast to nonlinear media that allow users to retrieve specifics items when they wish.
Longitudinal: type of survey designed to collect data over several points in time. See cross-sectional.
LOP (least objectionable program): popular theory of television audience behavior, attributed to Paul Klein, that argues that people primarily watch television for reasons unrelated to content, and they choose the least objectionable programs. See passive audience.
LPM (local peoplemeter): simply a peoplemeter installed to provide local market audience measurement.
Market segmentation: practice of dividing populations into smaller groups having similar characteristics or interests in order to market goods and services more precisely. See demographics.
Mean: measure of central tendency determined by adding across cases, and dividing that total by the number of cases. See average, median, mode.
Measure: procedure or device for quantifying objects (e.g., households, people) on variables of interest to the researcher.
Measurement: process of assigning numbers to objects according to some rule of assignment.
Measurement error: systematic bias or inaccuracy attributable to measurement procedures.
Mediametrie: private research company owned by French broadcasters and advertisers providing audience measurement services. See JIC, TRCC.
Median: measure of central tendency defined as that point in a distribution where half the cases have higher values, and half have lower values. See average, mean, mode.
Meter: measuring device used to record the on-off and channel tuning condition of a television set. See SIA, peoplemeter.
Metro area: core metropolitan counties of a market area as defined by a ratings service. Metros generally correspond to MSAs.
Metro rating: program or station rating based on the behavior of those who live in the metropolitan area of the market. See rating.
Metro share: program or station share based on the behavior of those who live in the metropolitan area of the market. See share.
Minimum reporting standard: number of listening or viewing mentions necessary for a station or program to be included in a ratings report.
Mode: measure of central tendency defined as the value in a distribution that occurs most frequently. See average, mean, median.
Mortality: problem of losing sample members over time, typically in longitudinal survey research.
MRC (Media Rating Council): U.S. industry organization responsible for accrediting the procedures used by ratings companies and monitoring the improvement of ratings methodologies.
MSA (Metropolitan Statistical Area): urban area designated by the Office of Management and Budget, often used by ratings companies to define their metro areas.
MSO (multiple system operator): company owning more than one cable system.
Multiset household: television household with more than one working television set.
Multi-stage sample: type of probability sample requiring more than one round of sampling. See cluster sample, probability sample.
NAB (National Association of Broadcasters): U.S. industry organization representing the interests of commercial broadcasters.
Narrowcasting: programming strategy in which a station or network schedules content of the same type or appealing to the same subset of the audience. See block programming.
NATPE (National Association of Television Program Executives): U.S. industry organization of media professionals responsible for television programming.
NCTA (National Cable and Telecommunications Association): industry organization representing the interests of the cable industry.
Net audience: See cume, reach.
Net weekly circulation: cume or unduplicated audience using a station or network in a week. See cume.
Network: organization that acquires or produces programming and distributes that programming, usually with national or regional advertising, to affiliated stations or cable systems.
Network analysis: approach to managing and visualizing data. It is useful in analyzing complex systems in which the component parts (referred to as nodes) are linked and interdependent. Because the units of analysis are not independent it uses specialized statistical procedures.
Nonlinear media: media systems that deliver content as it is requested. Digital video recorders (DRVs), video on demand (VOD), and websites that allow users to view or download media are nonlinear media. See linear media.
Nonprobability sample: kind of sample in which every member of the population does not have a known probability of selection into the sample. See convenience sample, purposive sample, quota sample.
Nonresponse: problem of failing to obtain information from each person originally drawn into the sample.
Nonresponse error: biases or inaccuracies in survey data that result from non-response. See nonresponse.
Normal distribution: kind of frequency distribution that, when graphed, forms a symmetrical, bell-shaped curve. Many statistical procedures are premised on the assumption that variables are normally distributed. See skew.
NPower: package of computer programs offered by Nielsen that affords clients access to its national peoplemeter database. It performs a wide range of customized analyses of gross and cumulative audience behavior.
NSI (Nielsen Station Index): reports local television market ratings in the United States.
NTI (Nielsen Television Index): reports national television network ratings in the United States.
Off-network programs: programs originally produced to air on a major broadcast network, now being sold in syndication.
O&O (owned & operated): broadcast station that is owned and operated by a major broadcast network.
Opportunistic market: buying and selling of network advertising time on short notice, as unforeseen developments (e.g., cancellation, schedule changes) create opportunities. See scatter market, upfront market.
Optimizers: any one of several computer programs used by large advertisers or ad agencies that take respondent-level peoplemeter data as input to create advertising schedules that optimize campaign reach while minimizing costs.
Overnights: label given to ratings, based on meters, that are available to clients the day after broadcast.
Oversample: deliberately drawing a sample larger than needed in-tab to compensate for nonresponse, or to intensively study some subset of the sample.
OzTAM: private Australian company owned by the major commercial broadcasters that provides television audience measurement. See JIC, TRCC.
Page rank: method for scoring websites developed by Google that ranks them according to the number and importance of their inbound hyperlinks.
Page views: number of Web pages viewed by users in a given time period.
Paid search advertising: popular form of advertising offered by search engines, in which advertisers bid for the right to place advertisements in response to certain search terms. Advertisers then pay on the basis on how many users click on their ads (i.e., pay-per-click). Keyword search and pay-for-placement are other terms describing this practice.
Panel: type of longitudinal survey design in which the same sample of individuals is studied over some period of time. For example, meters are placed in a panel of television households. See cross-sectional, longitudinal, trend analysis.
Passive audience: term given to viewers who are unselective about the content they watch. Passive audiences are thought to watch television out of habit, tuning to almost anything if a preferred show is unavailable. See active audience, LOP.
Pay cable: programming services provided by a cable system for a monthly fee above and beyond that required for basic cable. Pay cable may include any one of several “premium” services like HBO, Showtime, or The Disney Channel.
Pay-for-placement: see paid search, keyword search.
Pay-per-click: pricing model in which advertisers pay based on the number of users who click on their ad. See paid search.
Pay-per-impression: pricing model in which advertisers pay based on the number of impressions delivered by ad servers.
Peoplemeter: device that electronically records the on-off and channel tuning condition of a television set, and is capable of identifying viewers. If viewers must enter that information by button pressing, the meter is called active; if the meter requires no effort from viewers, it is called passive.
Periodicity: problem encountered in systematic sampling in which the sampling interval corresponds to some cyclical arrangement in the list.
Phone recall: type of survey in which a phone interviewer asks the respondent what they listened to or watched in the recent past, often the preceding day. See coincidental.
Placement interview: initial interview to secure the willingness of the respondent to keep a diary or receive a meter.
Platform: technological system for delivering content to users. Broadcast television, cable, the Internet and mobile devices are all referred to as platforms.
PMSA (Primary Metropolitan Statistical Area): urban area designated by the U.S. Office of Management and Budget that is often used in defining ratings areas.
Pocketpiece: common name given to Nielsen’s weekly national television ratings report.
Pod: group of commercials, often numbering six to eight, that are aired one after the other. An ad’s position in the pod is sometimes negotiated.
Population (or “universe”): total number of persons or households from which a sample is drawn. Membership in a population must be clearly defined, often by the geographic area in which a person lives.
Portable peoplemeter (PPM): metering system in which respondents wear a small device (e.g. wrist watch or page-like) that detects an inaudible code embedded in a broadcast signal. When the meter “hears” the code it ascribes the respondent to the audience.
Post-buy analysis: analysis conducted after a program runs. It could refer to (1) a financial analysis to determine whether the price paid for the program was appropriate, or (2) the analysis of ratings performance to determine whether the predicted rating was correct.
Power ratio: statistic that expresses the relationship between share of revenue and share of audience. Also called the conversion ratio, or home market share ratio.
PPDV (persons per diary value): number of persons in a population represented by a single diary kept by a member of a ratings sample. PPDV is used to project an audience. See projected audience.
Preempt: action, taken by an affiliate, in which programming fed by a network is replaced with programming scheduled by the station. Certain types of commercial time can also be “preempted” by advertisers willing to pay a premium for the spot.
Prime access (in local television): first hour of prime time, prior to network programming. Local stations generally schedule syndicated programs or local productions during prime access.
Prime time: television daypart often from 7 P.M. to 11 P.M. EST. Due to FCC regulations, U.S. broadcast networks typically provide prime time programming to affiliated stations from 8 P.M. to 11 P.M. EST.
Probability sample: kind of sample in which every member of the population has an equal or known chance of being selected into the sample. Sometimes called random samples, probability samples allow statistical inferences about the accuracy of sample estimates. See confidence interval, confidence level, sampling error.
Processing error: source of inaccuracies in ratings reports attributable to problems inherent in the mechanics of gathering and producing the data. See ascription, editing.
Program type: category of programming usually based on similarities in program content. Ratings companies often use standardized program types to report and summarize program audiences.
Projectable: quality describing a sample designed in such a way that audience projections may be made. See projected audience, probability sample.
Projected audience: total size of an audience estimated to exist in the population, based on sample information. See HPDV, PPDV, probability sample.
Psychographics: category of variable that draws distinctions among people on the basis of their psychological characteristics, including opinions, interests, and attitudes.
PTAR (Prime Time Access Rule): FCC regulation, effective from the 1970s through the early 1990s, limiting the amount of network programming that affiliates could carry during prime time, and preventing affiliates in the top 50 markets from airing off-network reruns during prime access.
PUR (persons using radio): term describing the total size of the radio audience at any point in time. See HUT, PUT.
Purposive sample: type of nonprobability sample, sometimes called a judgment sample, in which the researcher uses his or her knowledge of the population to “handpick” areas or groups of respondents for research.
PUT (persons using television): term describing the total size of the television audience, in persons, at any point in time. See HUT, PUR.
PVR (personal video recorder): another name for digital video recorders (DVRs), which has fallen into disuse.
Qualitative ratings: numerical summaries of the audience that not only describe how many watched or listened, but their reactions including enjoyment, interest, attentiveness, and information gained. See engagement.
Qualitative research: any systematic investigation of the audience that does not depend on measurement and quantification. Examples include focus groups and participant observation. Sometimes used to describe any nonratings research, even if quantification is involved, as in “qualitative ratings.”
Quota sample: type of nonprobability sample in which categories of respondents called quotas (e.g., males), are filled by interviewing respondents who are convenient. See nonprobability sample, probability sample.
RAB (Radio Advertising Bureau): U.S. industry organization formed to promote advertising on radio.
RADAR (Radio’s All Dimension Audience Research): Arbitron’s syndicated ratings service for U.S. radio network audiences.
Random digit dialing (RDD): in phone surveys, a technique for creating a probability sample by randomly generating phone numbers. By using this method, all numbers including unlisted have an equal chance of being called.
Random sample: See probability sample.
Rate card: list of how much a station will charge for its commercial spots. Rate cards are sometimes incorporated with ratings data in computer programs that manage station inventories.
Rate of response: percentage of those originally drawn into the sample who provide useable information. See in-tab.
Rating: in its simplest form, the percentage of persons or households tuned to a station, program, or daypart out of the total market population.
Ratings distortion: activity on the part of a broadcaster designed to alter the way audience members report their use of stations. See hypoing.
Reach: total number of unduplicated persons or households included in the audience of a station or a commercial campaign over some specified period of time. Sometimes expressed as a percentage of the total market population. See cume, frequency.
Recency theory: idea that an ad is most effective if it hits consumers when they are ready to buy. This places a premium on reach and timing, rather than frequency of exposure.
Recycling: extent to which listeners in one daypart also listen in another daypart. See audience duplication.
Relative standard error: means of comparing the amount of sampling error in ratings data to the size of different ratings. It is the ratio of the standard error to the rating itself. See sampling error.
Relative standard error thresholds: size of a rating needed to have a relative standard error of either 25% or 50%. Often published in market reports as a means of judging ratings accuracy. See relative standard error.
Reliability: extent to which a method of measurement yields consistent results over time.
Repeat viewing: extent to which the audience for one program is represented in the audience of other episodes of the series. See audience duplication.
Replication: study repeating the procedures of an early study to assess the stability of results. In audience measurement, replications involve drawing sub-samples from a parent sample to assess sampling error.
Respondent: sample member who provides information in response to questions.
Response error: inaccuracies in survey data attributable to the quality of responses, including lying, forgetting, or misinterpreting questions. See interviewer bias.
RFID (Radio Frequency Identification): tiny chip or tag, often embedded in products, that emits a signal that can be picked up by a scanner. RFIDs might offer a way to measure exposure to print media.
Rich media: type of Web advertisement that features more dynamic, eyecatching, content. Such advertisements often require Web connections with greater bandwidth to be effective.
ROI (return on investment): commercial benefits attributable to the money spent to generate those benefits. Marketers, for example, try to assess the sales attributable to advertising. See conversion rates, pay-per-click.
Rolling average: ratings level based on the average of several successive samples. As new sample data become available, the oldest sample is dropped from the average. A rolling average is less susceptible to sampling error. See bounce.
ROS (run of schedule): method of buying and scheduling advertisements in which the advertiser allows the station or network to run commercials at the best time that happens to be available.
Sample: subset of some population. See probability sample.
Sample balancing: see sample weighting.
Sample frame: list of some population from which a probability sample is actually drawn.
Sample weighting: practice of assigning different mathematical weights to various subsets of the in-tab sample in an effort to correct for different response rates among those subsets. Each weight is the ratio of the subset’s size in the population to its size in the sample.
Sampling distribution: hypothetical frequency distribution of sample statistics that would result from repeated samplings of some population.
Sampling error: inaccuracies in survey data attributable to “the luck of the draw” in creating a probability sample.
Sampling rate: ratio of sample size to population size.
Sampling unit: survey element (e.g., person or household), or aggregation of elements, considered for selection at some stage in the process of probability sampling.
Satellite radio: mode of radio signal transmission that uses satellites to send digital quality programming to individual subscriber receivers.
Scatter market: period of time, just in advance of a given quarter of the year, during which advertisers buy network time. See opportunistic market, upfront market.
Search engine: Web site specifically designed to help Internet users find specific pieces of information on the World Wide Web.
Segmentation: practice of dividing the total market into subsets, often related to the needs of a marketing plan or the programming preferences of the population. See target audience.
Server-centric measurement: approach to estimating audiences based on gathering information from server logs. See user-centric measurement.
Servers: powerful computers that host and provide services to multiple clients. These include the computers that (1) provide media content and services to Web users, (2) manage access to the Internet, and (3) serve ads to visitors across websites.
Sets-in-use: total number of sets turned on at a given point in time. As a measure of total audience size, it has become outdated since most households now have multiple sets. See HUT.
Share: in its simplest form, the percentage of persons or households tuned to a station or program out of all those using the medium at that time.
SIA (Storage Instantaneous Audimeter): later version of Nielsen’s original audimeter that allowed the company to retrieve electronically stored information over phone lines.
Simple random sample: one-stage probability sample in which every member of the population has an equal chance of selection. See probability sample.
Single source: term used to describe a data set that measures all the desired variables (e.g., metered media use across platforms, product purchases, demographic and lifestyle variables) across a single sample of respondents. See data fusion.
Skew: measure of the extent to which a frequency distribution departs from a normal, symmetrical shape. In common use, the extent to which some subset of population is disproportionately represented in the audience (e.g., “the audience skews old”).
SMSA (Standard Metropolitan Statistical Area): former governmental designation of an urban area, once used by ratings companies to define local market areas. See MSA.
SPI (Sample Performance Indicator): measure that reports the percentage of people who were asked to join the sample and are providing usable data. It is a conservative measure that reflects both response rates and day-to-day fault rates.
Spill: extent to which nonmarket stations are viewed by local audiences, or local stations are viewed by audiences outside the market.
Spin-off: programming strategy in which the characters or locations of a popular program are used to create another television series.
Spyware: program secretly placed on a user’s computer that sends data on surfing activities back to an advertiser or other business.
SRDS (Standard Rate and Data Service): service that publishes the station rate cards and other information useful in buying commercial time. See rate card.
Standard deviation: measure of the variability in a frequency distribution.
Standard error: standard deviation of a sampling distribution. It is the statistic used to make statements about the accuracy of estimates based on sample information. See confidence interval, confidence level, relative standard error.
Station rep: organization that represents local stations to national and regional advertisers, selling the station’s time and sometimes providing research information useful in programming.
Station total area: Nielsen term meaning the total geographic area upon which total station audience estimates are based. The total area may include counties outside the NSI area.
Statistical significance: point at which results from a sample deviate so far from what could happen by chance that they are thought to reflect real differences or phenomena in the population. By convention, significance levels are usually set at .05 or lower, meaning a result could happen by chance only 5 times in 100. See confidence level.
STB (set-top box): small, computer-like box that satellite and cable systems use to assemble digital input into programming displayed on a subscriber’s television set. When loaded with the appropriate software, STBs can record and report the channel switching behavior of the set, forming the basis of audience measurement.
Stratified sample: type of probability sample in which the population is organized into homogeneous subsets or strata, after which a predetermined number of respondents is randomly selected for each strata. Stratified sampling can reduce the sampling error associated with simple random samples.
Stripped programming: programming practice in which television shows are scheduled at the same time on 5 consecutive weekdays. Stations often strip syndicated programs.
Superstation: independent television station whose programming is widely carried on cable systems around the country.
Sweep: in television, a 4-week period of time during which ratings companies are collecting the audience information necessary to produce local market reports for November, February, May, and July.
Syndicated Network Television Association (SNTA): U.S. trade association that supports syndicators’ efforts to sell commercial time.
Syndication: selling a standardized product to many clients. A syndicated program is available to stations in many different markets. A syndicated ratings report is also sold to many users.
Systematic sample: kind of probability sample in which a set interval is applied to a list of the population to identify elements included into the sample (e.g., picking every 10th name).
TAM (Television Audience Measurement): acronym widely used around the world.
Target audience: any well-defined subset of the total audience that an advertiser wants to reach with a commercial campaign, or that a station wants to reach with a particular kind of programming.
TARP (Target Audience Ratings Points): gross ratings points attributable to a specific target audience. See GRP.
Television household (TVHH): common unit of analysis in ratings research, it is any household equipped with a working television set, excluding group quarters.
Terrestrial TV: television that is received directly from an over-the-air broadcast signal.
Theory: tentative explanation of how some phenomenon of interest works. Theories identify causes and effects, which make them amenable to testing and falsification.
Tiering: practice of marketing cable services to subscribers in groups or bundles of channels called tiers.
Time buyers: anyone who buys time from the electronic media for purposes of running commercial announcements.
Time period averages: size of a broadcast audience at an average point in time, within some specified period of time.
Time spent: any one of several metrics that report the amount of time users devote to media products or outlets, including time spent listening, time spent viewing, time spent per visit, time spent per page, etc.
Total audience: all those who tune to a program for at least 5 minutes. Essentially, it is the cumulative audience for a long program or miniseries.
TRCC (tripartite research company contract): A kind of industry-wide organization in which the firm letting the contract is itself an audience measurement company. It typically has tripartite ownership including the media, advertisers and agencies. They are not unlike JICs, except that they typically provide at least some of their own measurement services. See JIC, mediametrie.
Trend analysis: type of longitudinal survey design in which results from repeated independent samplings are compared over time.
TSL (time spent listening): cumulative measure of the average amount of time an audience spends listening to a station within a daypart.
Turnover: ratio of a station’s cumulative audience to its average quarter hour audience within a daypart.
TvB (Television Bureau of Advertising): U.S. industry organization formed to promote advertising on broadcast television.
TVQ: ratings system that assesses the familiarity and likability of personalities and programs.
UHF (ultra high frequency): class of television stations assigned to broadcast on channels 14 through 80.
Unduplicated audience: number of different persons or households in an audience over some specified period of time.
Unique visitors: unique Web users that visited a site over the course of the reporting period. See cume, unduplicated audience.
Unit of analysis: element or entity about which a researcher collects information. In ratings, the unit of analysis is usually a person or household.
Universe: see population.
Unweighted in-tab: actual number of individuals in different demographic groups who have returned usable information to the ratings company.
Unwired networks: organizations that acquire commercial time (usually in similar types of programming) from stations around the country and package that time for sale to advertisers.
Upfront market: period of time several months in advance of the new fall television season during which networks, barter syndicators, and major advertisers agree to the sale of large blocks of commercial time for the broadcast year.
URL (uniform resource locator): unique address for each document and resource on the World Wide Web. The system was developed in 1994 by Tim Berners-Lee.
User-centric measurement: approach to estimating audiences based on gathering information from a sample of willing respondents, often members of a panel. See server-centric measurement.
Validity: extent to which a method of measurement accurately quantifies the attribute it is supposed to measure.
Variable: any well-defined attribute or characteristic that varies from person to person, or thing to thing. See demographic.
VHF (very high frequency): class of U.S. television stations assigned to broadcast on channels 2 through 13.
View-through: measure of Web ad effectiveness that tracks how many Web users take action within some period of time (e.g., 30 days) after having seen an ad. See click-through.
Views: gross measure of the audience based on a count of the number of times a particular Web page or video has been seen.
VOD (video on demand): technology for delivering video (i.e., television programs, movies) in response to a viewer’s request. Cable systems and some websites support VOD services. See nonlinear media.
VPVH (viewers per viewing household): estimated number of people, usually by demographic category, in each household tuned to a particular source.
Web analytics: practice of collecting and analyzing the data produced by the servers that power the World Wide Web. Analysts often use software packages to study the behavior of visitors to a particular website, with the purpose of managing the site to the owner’s benefit. See server-centric measurement.
Web page: electronic document that can contain several files (e.g. text, pictures, links, etc.). Websites often contain many pages, each of which typically has its own URL. Pages form the basis of many Web metrics including pages views, time spent per page, etc. See URL.
Website: specific location on the World Wide Web offering information, entertainment, and/or advertising.
Weighted in-tab: number of individuals in different demographic groups who would have provided usable information if response rates were equivalent. See sample weighting.
Weighting: process of assigning mathematical weights in an attempt to correct overrepresentation or underrepresentation of some groups in the unweighted in-tab sample. See sample weighting.
WFA (World Federation of Advertisers): A trade association advocating for major companies engaged in global marketing.
WWW (World Wide Web): system of protocols and programs that enables Internet users to access pages of information on computer servers around the world. See URL.
Zapping: practice of using a remote control device to avoid commercials or program content by rapidly changing channels. Often used interchangeably with zipping.
Zipping: practice of using the fast-forward function on a VCR or DVR to speed through unwanted commercials or program content. Often used interchangeably with zapping.