Conservation research includes social as well as natural sciences, and social sciences include qualitative, qualitative, and mixed methods. Most mixed-methods research adopts quantitative and qualitative approaches in parallel or in sequence, using related but distinct sets of data. Transformations of raw data, to apply qualitative and quantitative approaches to the same dataset, are uncommon, and beset by obstacles. I argue that dual analysis of a single dataset can be valuable and sound, subject to four caveats. Authors must: check for adequate sampling as well as theoretical saturation; distinguish clearly between structural and statistical associations; explicitly describe conversion from qualitative constructs and codes, to quantitative categories or other variables; and ensure that precision, reliability and generalizability are considered appropriately within both analyses.