To analyse the dynamic patterns in framing within negotiation, some researchers have employed socio-linguistic and discourse analytic methods (e.g., Donnellon & Gray, 1989; Gray et al., 1997). These methods have involved inferring the meaning that disputants construct of their experience of the conflict by analysing their behaviour and the content of their expressions in systematic ways. Although these approaches have been satisfactory for understanding the way that frames operate to affect the negotiation process in small numbers of cases, they are nonetheless constrained by their limited generalisability across settings. To investigate the underlying processes in more detail, some negotiation researchers have been known to analyse the negotiation process in stages (Holmes, 1992; Lytle et al., 1999; Olekalns, Smith & Walsh, 1996).
To divide the negotiation process into segments for closer analysis, some investigators have drawn upon broader negotiation theory (Holmes, 1992). Prescriptive stage models argue that there are a minimum of three stages through which disputants proceed to reach a successful resolution (Holmes, 1992). These tripartite models typically propose that disputes consist of an initiation or an information exploration stage, a problem-solving stage, and a final resolution stage. More comprehensive models, such as Littlefield, Love, Peck and Wertheim's model (1993), prescribe an additional introductory stage wherein negotiators discuss what they want from the process. Other prescriptive problem-solution focussed models such as D'Zurilla's conflict resolution model (1988) contend that five stages are needed for successful resolution: problem orientation, problem definition and formulation, generation of alternative solutions, decision-making, and solution implementation and verification stages. However, when comparing prescriptive models to descriptive models of successful negotiation, there is limited empirical support for the more comprehensive stage models. Descriptive models indicate that successful conflict resolution follows a three-stage process and that the three stages largely resemble those detailed in the prescriptive tripartite models (Holmes, 1992). In keeping with these findings, the current research analysed the negotiation process across three stages as per the tripartite stage models. However, because the chief focus of this study was upon determining shifts in frame use across time, and not necessarily upon examining frame use according to stage theories of negotiation, the plan was also to assess frame use across two and four stages.
In analysing the negotiation process in stages or segments, this study investigated whether conflict frames shifted over time and in particular, determined whether frames converged, diverged, or were maintained over time. In assessing these patterns, frame convergence was defined as the movement of both negotiators to a more similar level of frame use in the later stage as compared with the early stage of the negotiation. Consistent with the communication-related definitions of convergence (Giles et al., 1991), this definition also incorporated instances wherein one party converges to the other in their frequency of frame use, and the other party maintains the same frequency of frame use. Divergence was defined as movement to a less similar frequency of frame use between negotiators in the latter stage as compared with the former stage of the negotiation. As with the definition of convergence, this definition also captured instances wherein one party diverges in their use of frames and the other party maintains the same frequency of frame use. Frame maintenance was defined as instances in which there is a similar frequency of frame use between negotiators in both the former and latter stages. To capture every possible pattern in the frequency of frame use between the two stages, a further category was identified and termed frame shift equivalence. This category refers to a pattern in which negotiators simultaneously increased or decreased their level of frame use to the same extent across the negotiation. Frame shift equivalence implies that there is no difference in the use of frames between negotiators at the one time, but that there is a difference in joint use between stages of the negotiation. This pattern was of exploratory interest in the current research.