Can we still trust data? Following a series of surprising election results in 2016 that defied the predictions made by some of the world’s most talented pollsters, some critics seem to think the answer is a resounding no, writes Daniel Castro in Government Technology. But while those who forecast elections certainly made mistakes, these failures do not mean that other uses of predictive analytics should be discarded. To the contrary, a close look at U.S. presidential election predictions shows that more investment in data, not less, is the way to avoid replicating these types of problems in the future.
