This chapter explains the link between Articles 2 and 3, i.e. how the third and last study builds on the findings and insights generated in the second study. This is followed by reflections on the choice of the experimental design and the elaboration of the manipulated texts applied in that design.
11.1. Explicatory Output from Article 2
Article 2 demonstrated that the participating companies operate with two separate value systems for corporate identity and corporate social responsibility, respectively, and that the CSR value system was heavily dominated by morally-based values suggestive of a traditional perception of CSR. In seeking to compare this result against the first study’s evidence of a new and different perception of CSR, the notion of frames seemed suited for the task of explaining these findings. Looking through the theoretical lens of frames and framing, there was clear indication that the corporate world operates with one dominant frame of CSR whereas consumers work with an almost contradictory or competing frame of CSR.
11.2. Generative Input for Article 3
Framing holds explanatory power as explained in Section 11.1., but it can also facilitate companies’
rhetorical creation of CSR messages based on clusters of perceptions and expectations of CSR for consumers and companies, respectively: the traditional CSR frame and the new CSR frame.
Rhetorical realizations of these frames by way of CSR messages provide the basis for exploring the third theme treated in this dissertation, namely the reception and interpretation of CSR messages.
11.3. Choice of Method
As discussed in section 6.3, the dissertation’s original research design suggested that focus groups would be effective in exploring consumer evaluations of differently framed CSR messages. This approach was replaced by a quantitative survey inspired by experimental design due to its capacity to test the impact of alternative framings of CSR messages on consumer perceptions of both companies and their communication. Focus groups would potentially generate richer data, but the danger of ending up with data that were negotiated meanings co-created in the focus group rather than evaluations of messages motivated the choice of the experimental design: The concern was that through discussions in the group, the participants would influence each other to an extent where they would change their opinions, or perhaps not feel comfortable in expressing their perceptions of the CSR messages (Halkier, 2008; Morgan, 1997).
11.3.1. An experimental design
Experimental designs are designs where the researcher manipulates the independent variable(s) in order to observe its effect(s) on the dependent variable(s). Hence, Shadish, Cook & Campbell (2002, p. 12) define an experiment as “a study in which an intervention is deliberately introduced to
106 observe its effects”. Canavos & Koutrouvelis (2009, p. 3) argue that process is central to experimental designs in that this design allows researchers to :
test the effectiveness of different strategies, thereby providing an opportunity for the improvement of the process through the reduction of common cause variation. That is, by understanding the current operation of the process and the factors that cause variation in the process outcomes, we can design an experiment by which we will understand more clearly the role that these and other potential factors play in the variation of process outputs.
This study investigates the process that is salient to consumers’ assessment of corporate communication of CSR. The different value combinations or frames in the texts are the factors that may cause variation in perceptions. The process output is the consumers’ perception of both company and communication measured in terms of credibility and relevance.
The final design does not, however, meticulously include all of the traditional components of an experimental design. The reason for deviating from the traditional experimental design is that I wanted to include a comparative dimension in the survey. By exposing all respondents to all of the four manipulated texts (the treatments), the survey allowed for the respondents to directly compare the four framings of CSR messages. Such comparison would not be possible by stringently following the traditional experimental design where each group of respondents only evaluates one text before the four evaluations are compared by the researcher.
11.3.2. Elaboration of the four framed texts
According to Hallahan (1999), it is possible to identify seven models or types of framing applicable to PR/ corporate communication: a) Situations, b) Attributes, c) Choices, d) Actions, e) Issues, f) Responsibility, and g) News. This study uses a combination of three framing types: Situations, Issues, and Responsibility. The Situations framing focuses on the relationships between individuals, which allows this framing to be applied as a tool to both produce and examine communication between corporation and consumer. The Issues framing serves to map and identify “alternative interpretations of social reality” (Hallahan, 1999), where the social reality to be interpreted is that of corporations’ role in society, including their responsibilities. The Responsibilities framing serves to capture causality and responsibility as well as legal and moral responsibility. Hallahan (1999, p.
224) argues that framing “is not merely useful but is essential to public relations “, and that “public relations professionals fundamentally operate as frame strategists, who strive to determine how situations, attributes, choices, actions, issues, and responsibility should be posed to achieve favorable outcomes for clients”. Hence, together the three framing types chosen provide the contextual and strategic foundation for creating the manipulated texts.
Resting on the contextual foundation, the rhetorical construction of the four texts was based on the different value-based frames identified in studies 1 and 2 (the personal, competence-oriented consumer frame of CSR, and the societal, morally-oriented corporate frame of CSR). The value-based framings of the texts were elaborated, tested, and refined a number of times. The texts used in the pretesting phase are included in Appendix 25. Testers belonging to the target group were asked to read the texts and indicate for each text whether they saw it to carry a) moral or competence
107 values, and b) socially- or personally-oriented values. After each test, the messages were refined so as to more precisely reflect the intended value combinations. Needless to say, to manipulate texts by way of the intangible concept of values is a highly qualitative task as our perceptions of values are fluid. However, the messages were elaborated until general consensus on their value contents was reached.
The Danish professional research and analysis agency, Epenion was employed in order to secure that the young consumers making up the final sample were both representative of the group as well as randomly selected. Appendix 26 contains the final survey, and Appendix 27 provides an overview of respondent characteristics. Appendix 28 contains the survey data material generated.
108
109