The Impact of News Use and News Content Characteristics on Political Knowledge and Participation

Abstract

Research has examined the relationship between traditional news media use and normatively important political outcomes such as knowledge and participation. However, most research fails to account for variations in the nature of news over time and across communities that could alter the fundamental relationship between exposure and these outcomes. Here two studies are presented—one with variation in news characteristics over time based on the American National Election Studies time series data and another with variation across local communities and newspapers within a single state during a single election year—to assess the hypothesis that the relationship between news use and political outcomes are moderated by natural variations in the nature of the news content and news outlets.

Notes

¹A number of studies have indicated a strong empirical relationship between this measure and knowledge measures that do allow coding responses on the basis of accuracy (e.g., Delli Carpini & Keeter, 1996; Eveland & Scheufele, 2000). Price (1999), in his review of various measures of political information, described the likes/dislikes measure as an “indirect means of assessing information” (p. 598).

²Fortunately, the correlation between these dichotomous exposure measures and the days-per-week exposure and attention measures in the years in which the ANES measured them are reasonable: r = .33 and r = .84 for newspaper days and attention, respectively, and r = .49 and r = .58 for television news days and attention, respectively. These findings suggest that, although we would clearly prefer to have more sophisticated measures of news use—at least days per week and preferably days per week and attention to the campaign in the given media form, the dichotomous measures appear to be roughly capturing the behavior of interest. It should be acknowledged that these self-reported measures of news use may be inflated (Prior, 2009), but the inflation should be consistent across years and data sets and therefore does not pose a serious threat to the validity of our results.

³A commercial research company conducted the survey between November 3 (the day after the election) and November 23, 2004. The response rate was 44% using AAPOR formula RR5. In the sample, women (55.7%) were represented slightly more than men. Age was measured continuously and ranged from 18 to 91 (M = 48.43, SD = 15.99). Income was measured on a 6-point ordinal scale from less than $15,000 to more than $100,000 (Mdn and mode = 3, $25,000–$49,999). Education was measured on a 7-point ordinal scale ranging from less than high school to a graduate degree (Mdn = 3, some college but not graduated, mode = 2, graduated high school). Newspaper use was assessed by asking participants how many days per week they read a newspaper (M = 4.30, SD = 2.82).

⁴Respondents indicated reading an out-of-state local newspaper from a town bordering the state under study were excluded from the analysis.

⁵Newspaper circulation figures were obtained from the 2004 Editor & Publisher International Yearbooks for both dailies and weeklies (Editor and Publisher International Yearbook, 2004a, 2004b).

⁶To validate the use of circulation of state newspapers in our data set as a sufficient measure of news quality, we gathered the number of awards each state newspaper had won in the year our survey data were collected. We used the Press Club of Cleveland Excellence in Journalism Awards 2005 database (http://www.pressclubcleveland.com). This awards database was selected because awards were given out for stories written in 2004, because all daily and nondaily state newspapers were eligible to participate, and because no other awards database (including those of the Ohio Newspaper Association, the American Society of Newspaper Editors, the Newspaper Association of America, and the National Newspaper Association) met both of these criteria. Awards covered numerous categories, of which we selected only those relevant to this study, including Election 2004 coverage, politics/government writing, best daily newspaper, best nondaily newspaper, and community/local coverage. Judges for the contest were out-of-state Press Club members, and awards were based on significance, clarity, writing, enterprise, effectiveness, originality, and other criteria.

Note. Coefficients representing hypothesis tests are shaded. Model 1: Level 2 N = 13, Level 1 N = 24,840; Model 2–5: Level 2 N = 13, Level 1 N = 15,688. ANES = American National Election Studies.

^# p < .10. *p < .05.

Note. Coefficients representing hypothesis tests are shaded. Model 1: Level 2 N = 10, Level 1 N = 18,421; Model 2–5: Level 2 N = 10, Level 1 N = 12,978. ANES = American National Election Studies.

^# p < .10. *p < .05.

⁷We have only a handful of election years (between 10 and 13) available to us due to a weak correspondence between the survey data and content data, producing low statistical power for H2 and RQ1. That is, our sample size for H2 and RQ1 are based on the sample size for the number of years. It is for this reason that we considered findings H2 and RQ1 to be significant with a lower threshold of p < .10. Nonetheless, we may still be overlooking meaningful effects due to our small level-2 sample size. Only time can eventually increase this sample size appreciably.

Note. Coefficients representing hypothesis tests are shaded. Model 1: Level 2 N = 13, Level 1 N = 24,840; Model 2–5: Level 2 N = 13, Level 1 N = 15,688. ANES = American National Election Studies.

*p < .05.

Note. Coefficients representing hypothesis tests are shaded. Model 1: Level 2 N = 10, Level 1 N = 18,421; Model 2–5: Level 2 N = 10, Level 1 N = 12,978. ANES = American National Election Studies.

*p < .05.

⁸All the statistics of Study 2 are available upon request from the authors.

Additional information

Notes on contributors

Yung-I Liu

Yung-I Liu (Ph.D., The Ohio State University, 2008) is an Assistant Professor in the Department of Communication at California State University, East Bay. Her research interests include civic engagement, media effects on different ethnic groups, and communication geography.

Fei Shen

Fei Shen (Ph.D., The Ohio State University, 2009) is an Assistant Professor in the Department of Media and Communication at the City University of Hong Kong. His research interests include media effects, public opinion, and the sociology of news.

William P. Eveland

William P. Eveland, Jr. (Ph.D., University of Wisconsin-Madison, 1997) is Professor of Communication and Political Science at The Ohio State University. His research examines the role of news and interpersonal communication in creating an informed and participatory citizenry.

Ivan Dylko

Ivan Dylko (Ph.D., The Ohio State University, 2011) is an Assistant Professor in the Department of Communication Studies at New Mexico State University and researches political implications of communication technologies.