In this section we have talked about correlation, and started to build some intuitions about inferential statistics, which is the major topic of the remaining chapters. For now, the main ideas are:
- We can measure relationships in data using things like correlation
- The correlations we measure can be produced by numerous things, so they are hard to to interpret
- Correlations can be produced by chance, so have the potential to be completely meaningless.
- However, we can create a model of exactly what chance can do. The model tells us whether chance is more or less likely to produce correlations of different sizes
- We can use the chance model to help us make decisions about our own data. We can compare the correlation we found in our data to the model, then ask whether or not chance could have or was likely to have produced our results.