The MID framework for the design of data-driven tools
Due to the adoption gap that exists between the computational outputs and the expected use, designing data-driven tools for learning and teaching does not end with generating computational outputs: it is critical to attend to the ways those outputs will be integrated into human practices. The information a computational system can provide shapes which educational practices are possible. To attend to these intertwined dependencies, it is crucial to engage expert practitioners to validate the interpretability (what) and utility (how) of the outputs. This methodology was applied to computationally generated exploration-support called formative fugues. It (1) helped identify the alignment of the fugues with the educational goals of the learning environment, (2) highlighted the interpretability of the information by identifying practices afforded by fugues, and (3) suggested recommendations for delivery and presentation of the fugues for the educators. This design technique is dubbed as Multistage Implementation Design (MID), is inspired from the implementation design technique by (Wise & Vytasek, 2017) used for evaluating learning technology designs.