How positive and negative sentiments influence the success of university-industry collaborations: a study of Knowledge Transfer Partnerships

This post has been contributed by Dr Federica Rossi, Dr Muthu de Silva, Dr Ainurul Rosli and Dr Nick Yip

While the factors that underpin the success and failure of university-industry collaborations have been investigated extensively, researchers in this field usually explore factors that are connected to the pre-existing characteristics of the collaborating individuals or organisations — including their level of motivation and absorptive capacity, and the quality of their human capital — or their strategic actions and management practices at the individual, organizational, and project levels. Instead, very little attention so far has been devoted to the influence of the collaborators’ own sentiments and perceptions on the success of the collaboration – in other words, whether the collaborators’ evaluation of the collaboration in terms of positive or negative sentiments, influences their perception of how beneficial the collaboration is, and hence its likelihood to continue over time.

Evidence shows that affective evaluations – which refer to an evaluative feeling states associated with the emotional quality – affect managerial behaviour, including the focus of managerial attention, decision-making and problem-solving strategies. Affective evaluations have also been found to have a broader influence on business outcomes, with recent studies suggesting that managers’ emotional displays may lead to better performance and that emotions influence workers’ pursuit of goals. Particularly in the context of inter-organizational collaborations, affective bonds have been linked to the development of trust, which increases the resilience of such collaborations.

In our paper, we develop a conceptual framework linking affective evaluations, perceived challenges and perceived benefits, and the likelihood of future collaborations, drawing from a diverse literature, including management and psychology. We then test this framework using sentiment analysis applied to 415 written reports, each of which describes the process and outcomes of a university-industry collaboration funded under the UK Knowledge Transfer Partnership scheme.

With this study we make important contributions of value to theory and practice on the role of subjectivity in the success of university-industry collaborations.

First, we show that the collaborators’ perceptions of challenges and benefits of the collaboration are significantly associated with the likelihood of repeated collaboration: collaborators who perceive their collaboration as being more beneficial are more likely to collaborate again in the future. We also find that affective evaluations play a role. Collaborations that are perceived as being more challenging tend to be associated with negative affective evaluations, which reduce the perceived benefits of the collaboration (and consequently the likelihood to collaborate again in the future). However, if some positive affective evaluation of the collaboration is also present, this has two important effects: it directly increases the perceived benefits of the collaboration and weakens the positive relationship between perceived challenges and negative affective evaluations. In other words, positive sentiments about the collaboration can increase the likelihood of future collaborations, even in situations that are perceived as being challenging.

Second, we showcase the sentiment analysis of textual data as a helpful foresight tool to identify those collaborations that are more likely to continue over time.

Besides generating theoretical implications, our findings are of significant value for practitioners, as they suggest that perceptions and affective evaluations should be managed to achieve successful university-industry collaborations. For example, the interim reports prepared by collaborators could be analysed using sentiment analysis tools to identify situations where participants experience negative affective evaluations. In such cases, prompt interventions that aim to alter the overall conditions of a given collaboration that give rise to negative affective evaluations, and to introduce conditions that could generate positive affective evaluations, could result in shifting perceptions of benefits and thereby indirectly affect the likelihood of future collaboration.

To read our full results, please see our study: De Silva, M., Rossi, F., Rosli, A. and N. Yip (2020) ‘Does affective evaluation matter for the success of university-industry collaborations? A sentiment analysis of university-industry collaborative project reports’, Technological Forecasting & Social Change,

The paper can be downloaded from this link: