Time-series designs and analyses

This article describes the way time-series designs are used in research on international negotiation and related processes. Both quantitative and qualitative applications are discussed. One use of the techniques is to predict known outcomes of historical cases of negotiation. Both inductive and deductive approaches have been used in studies that evaluate alternative models of the way that negotiators respond to each other through the course of the talks. Another use of the techniques is to evaluate the impacts of such interventions as mediations or combat (referred to as interruptions) on the dynamics of conflict between nations. A third approach involves probabilistic forecasting with Bayesian analysis. This consists of revising initial probabilities of events (coups, peace agreements) based on current information about indicators that signal the occurrence of the event. Qualitative techniques have also been used to capture changes in conflict processes over time. These include charting changes in typological categories or in the use of influence strategies used by national actors in enduring rivalries. They also include tracing of paths to agreement or stalemate in negotiation, documenting progress in small-group dialogues, and developing chains of communication leading to peace agreements. By combining several of these techniques an analyst can draw conclusions about the likelihood that an event will occur (Bayesian analysis), its impact on a process (interrupted time series), and the way it emerged from prior events (process tracing).