Cross-domain collaboration between data scientists and experimental scientists holds promise to accelerate discovery in science. However, past research revealed that the scientists face coordination challenges due to the differences in their workflows and habit of collaborating with each other. We argue that how scientists experience the value of information received may be one of the reasons behind troubles in coordination. Combining a survey, interviews, public and private sources, we collected the data on information exchanges among 38 scientists in the R&D consortium that aimed to advance early dementia diagnosis. Using exponential random graph models to test our hypotheses, we found that the scientists tended to see information provided by cross-domain ties as less useful, as opposed to matching-domain ties. Yet how the scientists experienced the value of information provided by cross-domain ties depended on the scientist’s domain. Data scientists were more likely to consider information provided by experimental scientists as useful than the other way around. Moreover, when data scientists and experimental scientists had the same number of cross-domain ties to each other, data scientists were more likely to see information provided by a greater number of those ties as useful, as opposed to experimental scientists. Our findings have theoretical implications for research on cross-domain collaboration and social networks as well as managerial implications for management of R&D consortia and transition to data-driven decision-making in science.
