The decisions behind choosing teammates for an interdisciplinary team are significant. Team assembly – the reasons behind individuals’ decisions about whom to work with in teams – likely play a key role in shaping crucial team processes, such as conflict and viability. This thesis advances a two dimensional taxonomy of team assembly where member decisions of who to team up with can be: (1) driven by team maintenance or task performance concerns (i.e., team versus task), and (2) based on individual characteristics or dyadic relationships (i.e., compositional versus relational). The effect of these four assembly mechanisms on resulting conflict and viability perceptions were tested in a sample of thirty-nine design teams enrolled in a master’s level human-computer interaction course (over three years). Within each of three cohorts, individuals self-assembled into project teams to develop a product that would better lives in some way. Relational team assembly was measured at week 1, compositional team assembly was measured at week 2, team conflict at week 5, 10 & 14, and team viability at week 14 using surveys. Hypotheses were tested using exponential random graph models to predict conflict tie formation based on dyadic assembly rules, and regression to test if relational team assembly mechanisms predict team viability. Results indicate that taskwork assembly mechanisms predict team conflict, but teamwork assembly mechanisms do not. Relational teamwork and taskwork assembly mechanisms do not predict team viability. Future directions of research in team conflict, team assembly, and team networks are discussed based on the current findings. This thesis contributes to science by providing an interdisciplinary model of team assembly mechanisms, and evaluates the model in explaining team conflict and viability.

This master's thesis is engaged in optimisation processes of PUQ (as for Purchasing Quality) department, Robert Bosch České Budějovice. This department is responsible for input control of parts for manufacture as well as check of new parts including their documentation and documentation updates. By optimisation we understand selection of the best variant from group of possibilities.