Compounding mixed-methods problems in frame analysis through comparative research

Abstract

This article presents a novel methodological program for frame analysis of textual data and discusses its application in cross-national research. The proposed approach systematizes recent theoretical developments in frame analyses, and offers a step-by-step program for frame identification and measurement, which synthesizes the implications of aforementioned developments into a coherent methodological framework. Frames are identified with analysis techniques borrowed from sociolinguistics, discourse analysis, and dictionary-free quantitative content analysis. The results of the identification process are subsequently validated via latent class analyses of data matrices, which are obtained with the help of coding techniques that utilize computer-assisted qualitative software packages. Since more interpretative, qualitative elements of frame analysis are combined with quantitative content analysis, findings are of high conceptual validity and generalizability. Comparative broadening of the approach considerably improves the quality of the detected frames.

Media Framings of the Issue of Turkish Accession to the Eu

“…An additional consideration that had to be taken into account in the identification of keywords was the different meanings attached to the same words in different languages and contexts (Koenig, 2006). In some cases, keywords appearing in one country or context would be virtually absent in another, or would be invested with importantly different meanings.…”

Using a Triangulated Theoretical Framework and Mixed Methods: Frame, Institutional and Network Analyses

“…The work of several authors informed the approach used here, albeit that not all of them focus on 'frame' analysis. The most instructive works came from: Koenig (2004Koenig ( , 2005Koenig ( , 2006, Benford and Snow (2000), Fischer (2003), Fischer and Forester (1993), Hajer (1995Hajer ( , 2006a, Hajer and Versteeg (2005), Hajer and Wagenaar (2003), Schön and Rein (1994) and several works by Dewulf, et al, 2009) Frame analysis is not a search for an objectively provable reality. Rather, it is an approach to seeking descriptions of the multiple conceptions that people have about particular issues.…”

“…Richards (2005) calls this data-theory bootstrapping. As patterns began to be found, rather than taking complete creative license to 'invent' another raft of master-frames, preexisting and tested master-frames were used as a guiding scaffold (Koenig, 2006). In this way, the frames and their differences were inductively derived from coding and comparing respondent statements related to the various issues of interest .…”

“…Frames are generally described as operating at two broad levels, although different authors define them slightly differently. The broadest level is often called a master-frame (Benford & Snow, 2000;Koenig, 2005Koenig, , 2006 or a meta-frame (Schön & Rein, 1994). The second level can be described as the diagnostic and prognostic frames (Benford & Snow, 2000), or as (Schön & Rein, 1994) (Snow, et al, 1986).…”

“…Working with closed-coding at the outset reduced the potential for an ill-defined exploration, and combining it with open-coding left space for change and discovery (Weston, et al, 2001:385). The first step in the more detailed coding looked for the argumentative schemes, the common normative claims and descriptions (Koenig, 2006) participants used to organise their views and construct meanings of fire management and of climate change within a policy context . This meant looking 'within' each argument for concepts and metaphors that participants drew upon to describe the causes of the policy problems, and how they linked these to conceptions of means to addressing the problems.…”