“Sources and Journalists” Revisited: Proposing an Interdependent Approach to Source Use

ABSTRACT

This study analyzed print journalists’ perceptions of source use during the 2014 U.S. midterm elections. Rank-order measures required journalists to compare five prevalent source types, revealing patterns in the acquisition of news content. While one or two sources (e.g., politicians and experts) exhibit a strong lead on some content, other source types are interchangeable, speaking to the “specialized” nature of certain information. Further analyses demonstrate a dichotomy, with patterns in acquiring traditional journalistic content (e.g., authoritative, credible, articulate) differing from those related to content defined as “soft” or “infotainment” (e.g., drama and emotion).

KEYWORDS:

• Midterm elections
• sourcing
• politics
• newspapers
• interdependence

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

1 Aside from a “Republican wave,” which crippled the Democratic Party, the election also featured several close races, including “the most expensive race in US history on a per-vote basis,” the continued rise of “dark money” in politics, and unprecedented use of social media by Republicans within a midterm election (Hendricks & Schill, 2016, p. 3). Thus, studying this midterm election is important in its own right, while also improving our understanding of U.S. midterm coverage generally.

2 While Hall et al. (1978) labeled those in power primary definers, journalists, who translate the information primary definers provide into news narratives, were labeled secondary definers.

3 See footnote 1.

4 In total, 80 journalists “completed” the survey, for a 32% response rate. However, two journalists answered “No” to the initial check question (said they did not cover the midterms) and were routed to the end of the survey.

5 If it were possible, these would have been included as controls in all analyses; however, due to the low sample size and the nature of the analyses, this would have resulted in too many zero cells.

6 6. For a complete list of questions, see online Appendix B.

7 Follow-up pairwise comparisons employed Bonferroni corrections.

8 Because at least three questions are normally needed to calculate scale reliability, Spearman's Rho was used.

9 The fact that “official” was included in the name of the source may have also resulted in a conflation of the two.

10 Analysis of the distribution of these outcomes showed that in some instances, the outcome for citizens was negatively skewed while the outcome for experts was positively skewed. Given this was not the case for all content types and the majority of outcomes were not skewed, I proceeded with repeated measures ANOVA. Results of ordinal logit regression (available from the author) revealed similar findings.

11 Analyses were replicated substituting questions from the second battery. These results are available from the author.

12 Disagreements exists as to the conceptualization and operationalization of hard vs. soft news; however, drama and emotion are consistently placed under the umbrella of soft news.

13 In the process of data collection, it became apparent that U.S. journalists are averse to taking surveys. Future research might systematically examine journalistic perceptions of survey participation and how this relates to organizational rules and journalistic norms.