The advent of big data creates new opportunities for errors in inference-making. Because the data is big, the errors can be similarly big (though the two notions may not be correlated.
In this piece for Wired Innovation Insights, an example is taken a study that found a correlation between Google searches for “southern cooking” and “passion live.” Did this correlation allude to an urban legend connecting southern cooking to libido enhancement — similar to the Snopes-reported belief that saltpeter would have the opposite effect. (Fried pickles credit: Will Folsom Flickr)
Big Data enablers like Teradata, Cloudera and Hortonworks allow analyst-friendly tools like PowerBI, Tableau, Spotfire and Qlikview to enter the workplace at desirable price points. Yet few would disagree that pairing easy to use tools with lightly trained analysts could have undesirable results ranging from unproductive labor to ill-advised recommendations.
Add to this heady mix Bring Your Own Device (BYOD) and rogue IT infrastructure facilitated by low cost cloud resources.