Predicting Image Inclusion and Image Type in Newspaper Homicide Reports Using Established Crime Reporting Themes

• In Event: Posters I

Abstract

Print news media attach images to articles to augment textual reports with a visual component. Public opinions surrounding crime can be greatly impacted by the content and themes presented in crime news, including associated images (Peelo, 2006). The current study identifies the most robust predictor of image inclusion among three predominant themes of crime reporting: sensational stories, victim relatability and fear of personal victimization (Pollak & Kubrin, 2007; Sacco, 1995). The analysis also examines the predictive capability for each of these themes in relation to different types of image content. A 12-year search of The Vancouver Sun newspaper archive (January 1, 2004-December 31, 2015) yielded close to 4,100 articles whose primary focus was reporting on a homicide that occurred in British Columbia. Nearly 40 percent of the articles contained at least one image; images were coded into categories involving people, scenes, and objects. Multivariate logistic and multinomial logistic regression models, accounting for each of the three primary themes of crime reporting, were used to determine how well each theme predicts both image inclusion and type of image included. Results are discussed in the context of news media outlets’ decision-making processes in relation to the use of photographs in homicide reports.

Authors

• Walter Works, Simon Fraser University

• Jennifer Wong, Simon Fraser University

• Jason Gravel, University of California, Irvine