|
"Comment: Media Use Measures for New Generations"
By Robert Y. Shapiro
With the ongoing changes in the American National Election Studies and with current methodological debates about how to measure mass media exposure and attention through self-reports, this is the right time to propose new media use measures for these important, high quality, and widely academic used surveys. Scott Althaus and David Tewksbury, among others, have thought hard about these measurement issues, as reflected in their clear-headed and persuasive report, for which we owe them much gratitude. Their recommendations and justifications are a step forward in resolving certain decisions on question wording and measurement. In getting to these recommendations Althaus and Tewksbury ask the right questions in the rights ways. They aptly cite the need to include the internet and radio, to focus on behavior during a “typical week,” and to continue to focus on the election campaign and to include measures of knowledge in the surveys.
But after reading their report, I am led to ask, should they go ever further? That is, to think further into the future of media use and its measurement? Is it now time to think ahead beyond the immediate mission of the next few ANES surveys? The title of the report refers to “a New Generation of Media Use Measure,” but why not ask about media use measures for new generations? This requires that we think into the next decades not just the next elections. What will the news media and media use look like thirty years from now? We have gone from the newspaper filled days of the 1950s, with their relatively sparse television news coverage, as televisions just became common in households and as radio was listened to widely, especially in people’s cars. Fast forward to the present that Althaus and Tewksbury examine, and then to future technology and media use patterns that will look more like the virtually paperless society of the Star Trek televisions series’ (where people might only possess a single book or two for sentimental value) than today’s information environment. And if this will be the case, should we not start looking ahead to a relatively paperless environment (even if there are people like me who still insist on printing out much of what they read)? What will journalism be like in this environment (and what will it mean to be a journalist)? These are daunting questions. And, as challenging as it may sound, should we just be designing surveys to study media use in the United States in isolation? Should our survey measures lend themselves directly to comparative study or to later adaptation to compare media use in the U.S. with that in other countries. This latter question was beyond the scope of Althaus and Tewksbury report, but it is surely a something that the ongoing Comparative Study of Electoral Systems (CSES) should be considering, ideally sooner rather than later.
But what can we do now to set the stage for studying change over time in media use in a more futuristic way. Althaus and Tewksbury’s work provides a good starting point, using the 2006 ANES pilot study. Is there more insight that we might get from this survey and from additional pilot studies that might be done, along with the major ongoing surveys conducted by the Pew Center for Research on People and the Press (e.g., its 2006 study), and the Project on Excellence in Journalism (PEJ; see its comprehensive latest report, The State of the New Media 2007: An Annual Report on American Journalism)?
To get a sense of possible changes over time these data should be explored further with regard to the likely effects of generational replacements. This assumes that age differences in responses to the survey instrument represent cohort, not aging effects. Given what we know about the use of new technology, the most plausible explanation for these differences is a cohort effect making generational replacement a major source of change and age differences a leading indicator into future media use. Further, with respect to the all of the analyses reported in the Althaus and Tewksbury report, it is important that the scales etc. developed yield the same results and conclusions across age groups.
Concretely—empirically—what else does this mean that Althaus and Tewksbury and others should be doing and perhaps advocating? For one, at some time soon, if not now, media use and exposure measures need to be refined to ascertain exposure to radio, television, and newspaper or print journalist as it occurs through the technology of the future, which would very likely be hand-held computer devices and the computers that would replace current television and radios at home, in automobiles, and virtually everywhere. This would allow media users to target information sources not just with the latest search engines but also just through the equivalent of channel searching or “surfing” involving the latest voice and touch screen technology, which would continue to allow for serendipitous or accidental exposure to new information. With this type of delivery of information—arguably virtually all delivered by computer—how do we distinguish (or do we even need to distinguish?) what we now call television, radio, newspapers, and the internet? We already have television, radio, and newspaper and other print sources delivered over the internet, and we need to ask to what extent and how do we parse out from this what would be defined as television, radio, and newspaper and other print sources? This assumes, of course, that these still originate from separate news outlets and organizations. At minimum, this would mean distinguishing responses about television, radio, and newspaper exposure that occur through computer connections from sources on the internet from which respondents are “watching,” “listening to,” or “reading” news online. Last, in developing such questions, the wordings and formats of the questions used to track trends over time should be constructed and tested so that they can be asked without consequence in different survey modes: face-to-face, telephone, and online (without “mode effects”).
What Althaus and Tewksbury and other media researches can also do to try to stay ahead of the curve in measuring validly and reliably media use and their effects is to encourage cognitive testing of measures and, as suggested above, to see if what appear to be valid and reliable measures continue to be so across new adult age cohorts. Althaus and Tewksbury’s report makes no reference to cognitive testing that went into the development of the 2006 pilot study. This warrants further attention now and in the future. The case that Althaus and Tewksbury make for the proposed 26 questions to measure information acquisition in “Appendix B” of the report is persuasive. As a new generation of measures—and especially as measures for new generations—their case would be more persuasive if they also could report that the measures in their best judgment appeared to have roughly comparable validity and reliability for both the newest and oldest age cohorts in the pilot study sample or in the samples that others have used to test the Decision Scale, the Closed-Mindedness Scale, and the Need for Cognition Scale. More difficult, however, is developing a measure or measures of political knowledge over time, which has become increasingly central to research on public opinion and voting and which was beyond the reach of Althaus’ and Tewksbury’s able report.
|
