In this research, we apply network analysis to data gathered through structured interviews and file forensics to evaluate the information flows and research workflows in materials science research projects. We discuss the gaps in reproducibility highlighted by this analysis and suggest practical steps to making these scientific workflows more robust and sharable.