A data set of 104 software development projects – an investigation on process ambidexterity and project performance

This survey data was collected from software development (SD) organizations based in the United States and Germany. The survey methodology allows us to collect a larger cross-sectional dataset from practitioners, which was required to test our mediation hypotheses. A research project description explaining the purpose and expected outcomes of the study was sent to 16 software firms through the network of an applied research institute. Ten firms agreed to participate in the study. A customized survey link was sent to a senior executive (e.g., CEO or CIO) in each participating firm via email. The survey was forwarded to a total of 364 SD project managers and members. To reduce the bias related to retrospective data, we defined valid respondents as those who completed at least one SD project in the respective firm within the past 12 months. After removing invalid and incomplete responses, we retained 214 usable responses from 104 SD projects, resulting in a response rate of 55.2%