-
For our final project in the #rstats programming course, we're going to try to build models from the NHANES dataset to predict three outcomes. Each outcome is a group and they have to coordinate so they each run different models. At the end, we'll compare results.
-
I know that this is probably not considered kosher, but it is a rich dataset and the point is for them to build predictive models and see how they do. It's not a contest, but the goal is to have collective insight into the structure of the NHANES data.
-
I've had good success with these kinds of collective projects in the past. I think it really highlights how different the analysis can be based on different kinds of reasoning about the data.
-
This week, their goal is to do EDA on the dataset, and decide on what predictors they'll use to build the models. We'll see what happens, but I think these kinds of collective projects are very helpful for students and they learn a lot from each other.
-
They've been primed to learn from each other from the tidyverse presentations, and the midterm projects, so hopefully this will be insightful for them.