Were you a mother who listened to classical music during your pregnancy, or a parent who played classical music in your newborn baby’s nursery because you heard it stimulates creativity and improves intelligence? If so, do you know where this “classical music makes you smarter” idea came from?
In 1993, a psychologist named Frances Rauscher, explained David Weinberger in his book Too Big to Know, played 10 minutes of a Mozart piano sonata to 36 college students and then tested their spatial reasoning. Rauscher also asked the students to take the same test after listening to 10 minutes of silence, and again after listening to 10 minutes of a person with a monotone speaking voice.
The results of this experiment seemed clear, reported Rauscher: “The students who had listened to the Mozart sonata scored significantly higher on the spatial temporal task.”
“This modest experiment,” Weinberger explained, “using a tiny, non-random group led to a small industry of Mozart for Babies CDs, Georgia’s distribution of free Mozart recordings to every newborn in the state, and even death threats against Rauscher for reporting that she did not observe the same beneficial results from rock and roll."
As Weinberger concluded, “thus a handful of data gave rise to conclusions far beyond its reach. Thousands of babies grew up listening to cloying New Age renditions of Mozart’s works because a statistically insignificant, non-representative sample of college kids did marginally better at a narrowly defined task under poorly controlled circumstances. That’s how science too often is taken up by our culture.”
Do big data grown-ups know better than Mozart Babies?
Of course nowadays, during the era of big data, we like to believe that we’re free from the folly of extrapolating weak correlations from small data sets. We like to imagine that we’re now better data scientists who are aware of the signal-to-noise ratio in big data. While we are getting better at using large data sets and predictive analytics, we are occasionally frustrated by things that we thought these advanced techniques would help us conquer, such as predicting the weather.
Despite our advances, as Tom Redman and I discussed, many individuals and organizations still struggle with fundamental concepts, such as if data quality matters less in larger data sets, if statistical outliers represent business insights or data quality issues, statistical sampling errors versus measurement calibration errors, and mistaking signal for noise (i.e., good data for bad data). All of which understandably makes many wonder whether the principles and practices of true data scientists will truly be embraced by an organization’s business leaders.
However, as Tom Redman cautioned, “the purpose of science is to discover fundamental truths about the universe. But we don’t run our businesses to discover fundamental truths. We run our businesses to serve a customer, gain marketplace advantage, or make money.” In other words, the commercial application of science often has more to do with commerce than it does with science. This is what Mozart for Babies teaches us about data science.
Keep this lesson in mind the next time you’re discussing the application of big data and data science to your organization’s business problems with members of executive management. Nevertheless, I guess it couldn’t hurt to try playing some classical music during your discussion.