Faced with ever-increasing pressure to innovate, firms consider universities to be significant sources of knowledge. Such knowledge flow can take place in a variety of ways such as academic publications and contract research, but also in more collaborative modes such as joint research projects. This paper focuses on a specific collaborative model, in which firms and universities are involved together in a PhD project. We analyze the effects of project management, communication, and supervision characteristics on the success of such PhD projects using a survey conducted at Eindhoven University of Technology, the Netherlands. We conclude that management decisions, supervision and communication characteristics have a significant impact on the ultimate success of a project. Among other things, the choice of university supervisor plays a pivotal role. Success is more likely if there is joint decision-making by both university and partner. We believe our findings help universities and firms to collaborate successfully.

A collaborative Ph.D. project, carried out by a doctoral candidate, is a type of collaboration between university and industry. Due to the importance of such projects, researchers have considered different ways to evaluate the success, with a focus on the outputs of these projects. However, what has been neglected is the other side of the coin—the inputs. The main aim of this study is to incorporate both the inputs and outputs of these projects into a more meaningful measure called efficiency. A ratio of the weighted sum of outputs over the weighted sum of inputs identifies the efficiency of a Ph.D. project. The weights of the inputs and outputs can be identified using a multi-criteria decision-making (MCDM) method. Data on inputs and outputs are collected from 51 Ph.D. candidates who graduated from Eindhoven University of Technology. The weights are identified using a new MCDM method called Best Worst Method (BWM). Because there may be differences in the opinion of Ph.D. candidates and supervisors on weighing the inputs and outputs, data for BWM are collected from both groups. It is interesting to see that there are differences in the level of efficiency from the two perspectives, because of the weight differences. Moreover, a comparison between the efficiency scores of these projects and their success scores reveals differences that may have significant implications. A sensitivity analysis divulges the most contributing inputs and outputs.