In a wide variety of studies over the past two decades, the media has consistently been identified as a significant social institution implicated in normalizing and disseminating anti-Muslim prejudice. Identifying and combating Islamophobic discourses has been a challenging proposition, however, due to difficulties in systematically identifying and evaluating “Islamophobia” within texts. Islamophobia is a complex and contested phenomenon that defies easy classification within the boundaries of terminology which would normally be employed to describe hostility or prejudice based on, for example, race or religion. This article outlines some of the challenges involved in defining and categorizing Islamophobic discourses through an exploration of the process of constructing, evaluating, and applying a unique content analysis instrument, the “Islamophobia Index,” to media texts, using data sets drawn from the Australian news media. We critically reflect upon the methodological limitations of systematic quantitative studies, the importance of qualitative and interpretive approaches that take into account researcher subjectivities, and ultimately reposition and repurpose this project as a mixed-method study. In conclusion, potential applications for the Index, including non-media textual analysis, are considered.
These data points were of particular interest as during the 2004 Federal Election campaign Islamophobic rhetoric deployed in political discourse was relatively unchallenged by mainstream Australian media, but the reporting on the 2007 Federal Election campaign featured more critical lenses being applied to such rhetoric, therefore providing key points of diverging discourses in political discourse worthy of deeper analysis (Anderson 2015).
Coders were recruited from an informal research network at an urban Australian university composed of early career researchers interested in exploring different ways of evaluating Islamophobia within a variety of texts. Coders were all from different disciplinary backgrounds (including International Relations, Human Geography, Linguistics, Peace and Conflict Studies, and Comparative Literature), with varying levels of experience researching matters relevant to the study of Islam and Muslim societies.
This is a departure from the rating scale utilized by Anderson (2015), which utilized a bipolar 9-point Likert style rating scale to assign descriptors ranging from “extremely closed” (aggregate Index score of –8 > –7) through to “extremely closed” (aggregate Index score of 7 > 8). The nine different descriptors used in this rating scale did not enable clear definitions of agreement on the degree of open and closed views in given units of analysis which made it more difficult to determine acceptable levels of statistical agreement between different coders, thus leading to our decision to utilize just three divisions for aggregate scores.