This book was a good reminder of how we can exploit data in a certain way to get an answer we are happy with. While we should not do this, it does occasionally get done by certain bad-faith actors who want their own biases to be the most well known. 

This book hits home on how we need to make sure that the data cane be replicated again and again to get similar or even the same results. Without that the data is not usually correct and something has gone wrong when looking at the data. While some of this could be due to human error, it is always good to double-check how data is being collected, and interpreted, and who is behind the data collection funding.