>This means I oftentimes find myself looking at the raw numbers of, say, gun deaths or immigration numbers, etc.
Data is meaningless without context and context is inevitably political and subject to bias.
Data is meaningless without context and context is inevitably political and subject to bias.
For example high levels of cholesterol are negatively correlated to heart infarction, mostly because the population at highest risk is under medication and has so extremely low cholesterol. In this case the data itself is meaningless as it does not contain its own significance. This does not mean that we can freely discard data we don't like, it is important and useful, just not necessarily intrinsically meaningful.
A good analysis is made so that it is obvious to justify this layer of interpretation.