| dc.description.abstract | Most journalists and editors agree that it is virtually impossible to produce and disseminate news that is completely unbiased, and even if they did succeed, each consumer s personal biases would still influence the message. News, in one form or another, is biased. On that the research community agrees. What the research community has yet to agree, is a succinct yet encompassing definition, a categorisation of its various forms, their manifestation, and their impact, and how users arrive at or make judgements of bias. The subjective and constantly changing nature of the construct means that agreement on these, or even what exactly bias is, may never be forthcoming. Despite this, there has been a broad range of research on bias in the dissemination of news in traditional mediums. Yet, bias in the dissemination of news online has yet to receive the same level of attention. The most obvious lacuna in the literature is the impact of the visual presentation of news websites on perceived bias in the news articles they contain.
This thesis begins with a review of major authors in the domain lamenting the lack of serious underpinning theoretical research. It then provides a comprehensive literature review, including detailed classifications of biases that impact the production and dissemination of news. Due to a lack of theoretical and empirical models and frameworks explaining how judgements of bias are made or how it should be measured, the overarching domain of credibility is explored which has an abundance of models, theories and frameworks explaining how users form judgements of credibility online. These are also well supported with established underpinning theories. As no empirical frameworks were identified to aid in the design, development and deployment of experiments to measure credibility or any of its individual dimensions and measures such as bias, a first of its kind empirical framework and accompanying experiment platform to support such experiments was created.
Bias has been shown to be a core dimension and measure of credibility, especially when users are judging news on the WWW. Research has also shown that the visual presentation of websites affects judgements of credibility. This research was instigated based on the supposition that judgements of bias in news articles are similarly impacted by their visual presentation. The first experiment, described in this thesis, demonstrated that perceived bias in a news article could be impacted by the inclusion or omission of individual features. The second experiment investigated the design characteristic of professionalism. It found that as the level of professionalism in the aesthetics of a news webpage decreased, the perception of bias in the news article it contained increased, and vice versa. This effect was common across all webpage/article combinations tested and articles ranging from relatively unbiased to extremely biased. The final experiment found that based on this knowledge, it is possible to predictively impact the perception of bias by adapting another visual characteristic of a news webpage, visual quality. This demonstrates that perceived bias in a news article is at least partially judged heuristically in a similar fashion to credibility of which it is a core dimension and measure. This is significant for both news website designers and news consumers. | en |
