We just finished up training a group of DePaul faculty and students this morning in the use of ATLAS.ti, an incredibly flexible and powerful tool for qualitative data analysis, and it occurred to me that it might be helpful to put together some resources that are useful for doing one type of (generally) qualitative research. Content analysis, a system for analyzing patterns in textual data, is a perfect method for evaluating structures in large samples of relatively uniform texts. However, many folks aren’t clear on exactly what content analysis is, or how it differs from just reading a group of texts and inferring meaning from them. Content analysis applies particular tightly framed questions to a clearly defined sample of texts.
My research (to this point) has used newspaper articles as its primary unit of analysis. Careful selection of your data is critical for an effective content analysis. Newspaper articles are great for content analysis for two reasons. First, they are relatively uniform. Comparability is crucial in this kind of textual analysis. Second, they have a relatively clearly defined (and well researched) social role. It’s easy to explain why these texts matter socially, and that they play a particular discursive function. Of course, there are a plethora of other kinds of texts that share these features.
Ann Morning’s “Reconstructing Race in Science and Society: Biology Textbooks, 1952-2002” (pdf) uses biology textbooks to assess how constructions of race have shifted over time. The uniformity of the format of textbooks through time enables a high level of precision in her comparison. Wheelock and Hartman’s “Midnight Basketball and the 1994 Crime Bill Debates: The Operation of a Racial Code“, uses the transcripts of congressional debates to assess how subtle racial codes influence policy. When doing content analysis, any particular set of texts can make a good sample, so long as they are relatively uniform and occupy a specifiable place in a discourse.
While textual analysis tends to be associated primarily with qualitative research, and much good content analysis is qualitative, the method lends itself particularly well to the quantitative analysis of ostensibly qualitative data. Recent research by Stephanie Mudge traces historical shifts in the alignment of political parties with neoliberal policies through a secondary analysis of data on party platforms. Other scholars analyzed media coverage of recent riots in France to assess how a variety of news organizations framed the causes and consequences of the 2005 uprisings in the suburbs of Paris.
Content analysis, in sum, is appropriate for a wide range of data sets and research questions. Its strength primarily lies in its ability to reveal some of the underlying structures of communication embedded within discourses, and in its ability to reveal some of the social processes that structure the kinds of discourses that circulate publicly. For any researcher with an interest in culture, communications, or politics, this brand of textual analysis is worth a try.