Abstract:

R&D and innovation studies often involve complex and highly uncertain tasks, which require experts to work together. This paper aims to investigate how collaboration affects the probability of success for an R&D (Research and Development) project while controlling some other factors or groups. Retrospective data of the projects managed in an R&D Centre of the Scientific and Technological Research Council of Turkey (TUBITAK) are used in the study. A generalized linear model, logit, and mixed-effect logistic regression in order to examine random effects, are implemented for the empirical analyses. In our findings, exceeding the project deadline appeared as a predictor of success in the R&D project, and collaborations’ effects on R&D project success are dependent on the type of these projects. It is useful to decide on the number of collaborating institutions, depending on the project type, type of funding, and the aim of the R&D projects. Product development projects aiming at digital government or homeland security will increase their probability of success via collaboration. R&D projects with limited funds, probably have concerns about the extra costs, but as odds ratios increase against expectations, we can conclude that these projects may also benefit from collaboration.

Keywords: