Pollsters and Journalists vs. Big Data
This article is of value because it challenges some of the ways in which we make our decisions. We are more persuaded by personal story than by data. We like a good story over a complicated story. We project our own preferences onto how we read data.
While we can never be without bias we can be aware of some of the bias’ we hold which, hopefully, will lead to us becoming careful researchers.
Don’t miss the comments section which points out some of the limitations of contemporary political polling.
Polls are descriptive not prescriptive so there is always a chance that people will choose to do something different than they say they will do. Thus, it was possible that I would be explaining why the big data pollsters got it wrong. But they didn’t. In fact, they were frighteningly accurate:
The take away for research and data analysis is that the more data one has, from more different sources, collected in different ways that point the same direction the greater confidence you can have in the analysis.