Researchers have typically adopted one of two temporal approaches to measuring conflict frames. Whereas some researchers have chosen to measure conflict frames before and after the negotiation (Pinkley & Northcraft, 1994), others have used a more interactive approach to the analysis of frames and have measured their adoption and re-adoption as the process unfolds (Donnellon & Gray, 1989; Gray et al., 1997). Although both measurement approaches assume that conflict frames are dynamic constructs that can change during the negotiation, the second approach acknowledges the unique information that analysis of the actual interaction can provide (Donnellon & Gray, 1989; Gray et al., 1997; Brett, Shapiro & Lytle, 1998). Moreover, the second approach enables greater insight into whether the adoption of particular frames at certain points in the process underpins the attainment of agreements during the course of the negotiation and resulting integrative outcomes (Drake & Donohue, 1996). The types of frames obtained from the two approaches may be different; the first approach is more focused on the information that negotiators provide about their expectations for and their reflections on the process, whereas the second approach provides information about how negotiators are experiencing the process as it occurs.
Researchers such as Donnellon and Gray (1989) and Gray et al. (1990) measured frames as expressed in communication during negotiation. To achieve this objective, each negotiation was first segmented into phases according to the nature of the issues in contention. Negotiator utterances within each of these phases were then categorised according to frame type.
The approach of Pinkley (1990; Pinkley & Northcraft, 1994) parallels the discourse analytic approach in that they also sought to characterise the essence inherent in descriptions of conflict. Pinkley's (1990) approach differed, however, in that the units of analysis were the pre- and post-descriptions of conflict and not phases of ongoing negotiator communication. This approach also differed in that each description was assigned a set of three frame scores depending on their location on each of the three frame dimensions.
Because the current research sought to analyse the use of frames as they occurred and, more specifically, to detect changes in frame use during negotiation, the most appropriate unit of analysis was the speaking turns taken by the negotiators. Not only did this approach enable assessment of frame use at a fine-grained level, but also enabled a more quantitative characterisation than would be afforded by discourse analysis.