1. Meals on Wheels, a program which delivers meals to the homes of senior citizens and people with disabilities, has been in the news recently.Interestingly, it is much better studied and understood than many federal programs.What sort of data would help you understand whether meals on wheels was successful, and what sort of statistical test or technique would you use to analyze that data?

1. In his book Superforecasting: The Art and Science of Prediction, Phillip E. Tetlock concludes that the people and organizations who are most effective at predicting future trends and events have several traits in common, including a habit of considering issues from many perspectives and highlighting the uncertainty inherent in their beliefs and conclusions.In what ways is this like (or unlike) the statistical approaches we have learned in this class?

1. We spent a lot of time in this course considering many different statistical tools, and when to use which tool.Give an example of a time when a choice of statistical tool made a meaningful difference in a result, and why we should care what tool is used for a given task.

1. Overfitting a model is a case where we pick a story (e.g. a line or curve) that fits the data too closely.Is overfitting an important concern?Why do you believe what you do?