Context: Collaboration depends on communication and upon having a similar understanding of the notions that are being discussed, and a similar appraisal of their value. Existing work seems to show that the collaboration between industry and academia is hampered by a difference in values. In particular, academic work focuses more on generalizing on the basis of existing evidence, while industry prefers to particularize conclusions to individual cases. This has lead to the conclusion that industry values scientific evidence less than academia does. 

Objective: This paper seeks to re-evaluate that conclusion, and investigate if industry and academia share a definition of scientific evidence. If evidence can be found of competing views, we propose a more finely grained model of empirical evidence and its role in building software engineering knowledge. Moreover, we seek to determine if a more nuanced look the notion of scientific evidence has an influence on how academics and industry practitioners perceive that notion. 

Method: We have developed a model of key concepts related to understanding empirical evidence in software engineering. An initial validation has been conducted, consisting of a survey of master students, to determine if competing views of evidence exist at that level. The model will be validated by further literature study and semistructured interviews with industry practitioners. 

Results: We propose a model of empirical evidence in software engineering, and initial validation of that model by means of a survey. The results of the survey indicate that conflicting opinions already exist in the student body regarding the notion of evidence, how trustworthy different sources of evidence and knowledge are, and which sources of evidence and types of evidence are more appropriate in various situations.

Conclusion: Rather than a difference in how industry and academia value scientific evidence, we see evidence of misunderstanding, of different notions of what constitutes scientific evidence and what strength of evidence is required to achieve specific goals. We propose a model of empirical evidence, to provide a better understanding of what is required in various situations and a better platform for communication between industry and academia.