|
|
|
Previously the low cost method was telephone interviewing, but since nothing could be shown to respondents, the key measure of attention getting value had to be recall. The evidence cited on the recognition page shows recall isn't nearly as accurate and complete. It all can be done on malls but that is one of the more expensive methods. Finally, all who are to be interviewed can be contacted at virtually the same time. The ease of replying at their earliest convenience generates replies faster, and in a form for producing instant tabulations. Results from BRC's pioneering experimentation with virtually all possible approaches to online sampling have been have been the subject of a number of articles and talks. They constitute an important part of the industry's validation of this new approach. In every case we were conducting parallel studies, comparing results from the types of surveys we have been conducting for years, with the results obtained by inviting people whose online address had been obtained in a variety of ways. This included online samples recruited by random-digit dialing to produce projectable online samples, panels recruited by giving both internet users and non-users web-TV, making it possible to survey both segments of the public online, balanced panels with demographics matching the total population and an expressed interest in taking surveys, as well as various samples drawn from everyone who left their e-mail address at a web site without indicating they did not want to be contacted. The first was a talk Don Bruzzone and Paul Shellenberg gave at the Annual Convention of the Advertising Research Foundation (ARF) in New York, March 2000. They showed the evidence we have compiled from two years of parallel testing of Super Bowl commercials online, and by our conventional recognition-based methods. The talk was published later in Quirk's July 2000. Results from the third year of parallel testing were presented before the ARF in October 2001, and from the fourth year in June 2002 at the IIR Conference on Web-based Surveys in San Francisco. The general result was that all forms of online research had a strong tendency to pick the same winners and losers as our conventional surveys. Further details are in BRC's "Top 10 Insights" from those conferences in the What's New section, reachable by clicking the button for that section that follows. BRC's interest in conducting research on its research continued in 2008 with a key experiment conducted as part of its annual online post testing of all Super Bowl commercials. In conjunction with Survey Sampling International and the ARF Council on the Quality of Online Samples two parallel surveys were conducted. One used a conventional online sample with its typically large number of frequent survey takers. The other was made up of people who had recently been recruited to take surveys by SSI but had not been asked to take one yet. It was set up to provide conclusive evidence on the effect frequent survey takers have on advertising research.
|