Abstract
Powerless language influences how media consumers perceive the believability and trustworthiness of news sources. Past studies have suggested females use more powerless language, which can negatively influence credibility. However, previous research also suggests powerless language may be preferred in certain contexts. To investigate powerless language use in written health media, a content analysis investigated health-related articles in issues of 12 magazines between October 2008 and September 2009. Analysis compared differences in powerless language frequencies relative to three variables: biological sex of the author, biological sex of the audience, and the magazine's overall topic focus. Female authors and health-focused magazines used more powerless language than male authors and generic-focused magazines, and powerless language was more often directed towards a female audience than a male audience.
Acknowledgments
This study is a part of Ms. Fandrich's master's thesis under the direction of Dr. Beck.
Notes
χ2 (3, N = 6,883) = 41.998, p < .001, V = .078.
Note. Percentages with no subscript in common differ at p < .05 using Holm's Sequential Bonferroni post hoc comparisons.
χ2(3, N = 6,883) = 91.145, p < .001, V = .115.
Note. Percentages with no subscript in common differ at p < .05 using Holm's Sequential Bonferroni post hoc comparisons.
χ2(3, N = 7,026) = 57.770, p < .001, V = .091.
Note. Percentages with no subscript in common differ at p < .05 using Holm's Sequential Bonferroni post hoc comparisons.
Coefficient of reliability = 2M/(N 1 + N 2) where M = number of units coded the same, and N = number of total units for each coder. The coefficient of .903 is considered excellent and is consistent with other content analyses (e.g., Grob et al., Citation1997).
Instead of coding data in two steps (first, code for powerful vs. powerless, then code for specific powerless language types), both steps were combined into one. Research training found this to be an acceptable approach. Thus, the reliability score indicates the coding decision of units into one of four categories: powerful language; powerless language: hedge; powerless language: hesitation; powerless language: tag questions.