Today we discussed various properties of datasets, and what one can do with them. There are different types of datasets, e.g. straightforward numbers as in surveys, or the behavior of celestial objects over time, or the results of medical research (control, test, placebo groups). Datasets can be characterized by the mean, for example, but this is not always an accurate representation of the global dataset. One has to analyze what the information is meant to convey before extracting certain parameters from the data. One powerful tool is correlation analysis. However, there are pitfalls that especially non-mathematically-trained people often succumb to, and that is seeing correlations where there are none. It's very important to understand the subtleties of correlation analysis.