In Today's World Of Big Data, Maybe The Data Are Letting Us Down

In an age of sophisticated computer networks and algorithms, advanced societies have come to depend on the great amounts of data at our fingertips. The question is, how dependable are the data. For instance, in Kansas City they tracked 12 months of rain forecasting to find the computer's batting average was no better than the Farmers Almanac old standby: "Rain before 7 ends by 11."

Consider these five sets of stats and see what -- if anything -- they tell us:

* Donors representing only 0.01 % of the US population contributed a staggering 28% of the $ 6 billion spent on the 2012 elections. [See Bloomberg.com] Conclusion? These donors are admirably generous OR cunningly manipulative. Same data, different conclusions!

* The percentage of children living in poverty in the US jumped from 19 to 23 in just the last 5 years. [Associated Press] Conclusion? The administration is coddling the poor OR not doing enough for the poor. Again, same numbers can lead to opposite conclusions!

* As the US pulls out of Afghanistan they will be destroying $7 billion worth of military hardware that will cost more to ship back home. [The Washington Post] Conclusion? The military is inefficient OR war is inefficient. Play with these numbers anyway you wish and you still end up crying!

* Americans own almost 300 million firearms, or nine guns for every 10 people. [Slate.com] Conclusion? We are the world's largest gun-crazy nation OR we know how to defend ourselves against the dangers of government!

* The upcoming birth of William and Kate's child will bump the UK economy by almost $400 million. [Los Angeles Times] Conclusion? The Brits love royalty OR the Brits love sex!

Are these examples too flippant for your empirical training? Probably so. Still, they among others should give us pause in our computational rush into a Brave New World. And that's one conclusion even the statisticians can agree upon!

Filed under: Uncategorized

Leave a comment