Statistical Modelling for Big and Little Data

While the difference between "Data Science" and "Statistics" disciplines is, at best, blurred, many people associate machine learning methods and big data with the former, and modelling and inference for small samples (little data) with the latter. We present a big data application where no sophisticated method at all is needed, a small data application where a partial modelling approach seems useful, and a big-and-little data application where we can borrow strength from limited information in a large sample, to improve estimation based on more detailed data in a small sample.