Capability Assessments - An Experimental Study of Capability Assessments with Multi-Actor Dependencies

Abstract: This thesis investigates how multi-actor dependencies affect the ability to assess capability. It is also investigating if the accuracy of capability assessments for tasks where multi-actor dependencies are present, depends on if capability is assessed individually or with all participating actors. Two multi-actor dependency models were considered representative for all cases. Dependency I represents a scenario where two actors have a common goal, but each actor’s individual performance is not affected by the other actor’s performance. Dependency II represents a scenario where two actors have a common goal and the first actor’s individual performance affect the second actor’s performance. It was concluded that the participants tended to be more confident when assessing multi-actor dependency tasks in pairs than individually, although there is no significant difference in accuracy between the assessments. A qualitative analysis suggests that individual capability assessments are more accurate for individual tasks and for Dependency II tasks than for Dependency I tasks. For assessment made with all participating actors there is no significant difference between the dependencies, although assessments for Dependency II tasks tended to be more accurate than Dependency I tasks.