We report on an exploratory effort to design an interest-based learning experience for high school (step) dancers to engage with concepts in mathematics and data science. We hypothesized that generating and analyzing data from their own dance movement, through motion and audio capture, would: (1) enable learners to form analogies between off-line embodied experiences and new abstract concepts; and (2) support motivation to learn due to perceived relevance and usefulness of data science to dance practice. Based on initial interviews to understand the specific needs and interests of the steppers, we developed some early prototypes for visual and acoustic analysis, concentrating on pose precision, tempo, and spectral characteristics (timbre) for. Teacher and student reactions to the tools demonstrated support for our hypotheses that off-line embodied cognition would help with new knowledge acquisition and that the perceived usefulness of data science would motivate learning. Several promising future directions remain to develop an interest-based and embodied data science curriculum.