|
Abstract:
|
Face-to-face meetings are at the heart of collaboration. The POWER structure (Vance, 2018) is an effective method to make these meetings efficient, informative, and productive. POWER stands for Prepare, Open, Work, End, and Reflect, five key steps that help ensure the domain expert’s questions are answers accurately and quickly. On another note, when students are learning statistics, the environment is a sterile system, where problems are typically completely solvable and optimal solutions are well known and achievable. However, when collaborating, data is often messy, timelines are short, and there are different methods to approach the problem. Depending on the situation, different levels of accuracy are expected. While a Bayesian Hierarchical Spatial model may theoretically produce more accurate results, a simpler Generalized Linear Model can offer a sufficiently accurate solution at a considerably smaller cost. Predictive models often have a bias-variance trade-off, while collaborations have a usability-accuracy trade-off.
|
