This study compares gender differences on Likert scale pre/post assessments of engineering interest, identity, and knowledge in three " traditional" introductory function- and task-oriented robotics courses and two biomedical robotics courses. In addition, the STEM Academy at a local high school is surveyed to identify their preferences given six hypothetical robotics curricula: three traditional function- and task-oriented courses and three contextualized courses consistent with helping society and gender-friendly messaging. The students are asked to rate each hypothetical course from least to most preferred. ANOVA is used to test our hypothesis that the biomedical robotics curriculum will result in higher gains in engineering interest and identity for all students, especially for the girls in the sample. This study adds to the literature base by empirically testing the role that the design problems and contexts we choose as engineering educators plays in the gender inclusiveness of K-12 engineering education efforts.