Abstract
When conducting an architecture evaluation it is important that the right persons participate, so that as many views and aspects as possible of the architecture candidates are examined before development begins. At the same time, it is not cost efficient to include all stakeholders that might possibly have an opinion about the system. In this paper we investigate the amount of agreement between participants in an architecture assessment. The purpose of this is to identify which participants will provide unique views during a discussion and which participants share a similar view.