A company called ASI (now lpsos-ASJ) first used auditorium testing in the evaluation of TV programs, commercials, and movies in the 1960s. They used what they called the "Preview House" in Los Angeles as a controlled environment for such tests. Forty years later, versions of auditorium testing are still used by numerous market research firms around the world.

    The basic methodology in auditorium testing starts with a careful consideration of the goals to be achieved. The client advertiser, radio station, or TV station needs to identify, in the most precise way possible, the boundaries of the testing and how the results will be used. Once the desired outcomes are known, the researchers design a testing strategy to achieve those outcomes.

     With the strategy set, the research company screens and selects a group of between 75 and 200 people reflecting the demographic the client wishes to study. That demographic (a grouping according to age, gender, income, etc.) can be a random sample or one that is consistent with the station's current or desired audience, or even a subset of the audience that the client wishes to cultivate.

     The assembled test group is then invited into a small auditorium and given instructions for the test. They are rarely told what is being tested or who the client is for the testing. In fact, tests often include decoy selections to keep the participants from guessing which specific radio station or product is being evaluated. The test subjects are instructed to respond to samples of music, voice, messages, images, or other content, providing some sort of rating on a scale created by the researchers. This can be accomplished with written questionnaires, a joystick-type device that measures responses electronically, or even with a show of hands. Sometimes anecdotal comments are also solicited. Participants in auditorium testing are usually compensated for their time in order to increase the seriousness with which they approach the evaluating. The results are then tabulated and evaluated, with many variables charted, and correlations are made among the different samples tested. Ultimately the research firm can provide clients with both a review of the raw data and recommendations on how they may proceed to achieve their goals. Auditorium testing is essentially a hybrid of several market research methods, taking the group dynamic of focus groups, the larger size of diary or phone research, and the immediacy of one-on-one surveying.

     Bob Goode developed a form of auditorium testing called Electronic Attitude Response System. This method uses a video readout of averaged responses of the participants correlated directly to the audio content being rated and allows researchers to determine the test audience's preferences along with their "tune-out" of program elements. It also provides researchers with a sense of which program elements are more effective if paired with others. For example, a commercial following a weather report may lead to less tune-out than that same commercial aired after a musical selection.

     Music testing is a particular strength of auditorium testing models. Whole pieces within a musical genre can be tested before they are aired on a station. More commonly, however, "hooks," short segments of songs, are tested. In markets where many stations compete for listeners within each programming genre (country, oldies, urban contemporary, etc.), the subtleties of which songs are most liked within each genre can make a major difference in the ratings successes of each station. One firm, The New Research Group, offers 600 to 1,200 musical hooks along with 100 perceptual questions, allowing the client to know not only which music is preferred, but why, in specific descriptive terms, dealing with emotions, motivations, associations, etc.

     Research firms "cluster" music that appeals to test audiences in auditorium groups, because people who enjoy one song from the cluster are likely to enjoy others as well. In addition, firms use complex matrix charts to show compatibility between clusters, showing radio programmers how to broaden appeal by including more musical selections without causing tune-out by core listeners.

     Auditorium testing, along with other music testing, is seen by some as limiting, in that the short hooks it tests can over­ simplify otherwise interesting music that might gain acceptance upon being heard by audiences. For example, a hook from "Hey Jude" by the Beatles might not have tested well, whereas the song in its entirety was a number one hit.

     Public and commercial radio stations use research, including auditorium testing, to make program decisions; for example, the Wisconsin Public Radio network has been involved in the Corporation for Public Broadcasting program research, and Denver-based Paragon Research studied public radio stations in eight markets using focus groups, surveys, and auditorium research.

     As audio broadens its reach through new technologies such as satellite, Web-casting, and other distribution channels, it is likely that increasing specialization of program channels will occur, making auditorium testing more important in the precise selection of program content.

     That being said, the increasing sophistication of audience behavior measurement technology embedded in some of these new communications media may eventually render traditional auditorium testing too slow and imprecise by comparison for the emerging information needs of the industry.