In the Future, Everything Will Be In Standard Deviations
May 15, 2012 Leave a comment
Kevin Carey’s jury-duty-inspired rant about the need for more focus on statistics education is spot on — from an “allows you to contribute to society” standpoint, few things are more versatile. But it’s also worth noting that there are a lot of potential collective benefits to having a population that is statistically literate.
For example, there is enormous ambiguity and inefficiency in the way we describe original experiences. When somebody is asked what they thought of a new restaurant they’ll generally respond with “nice, “good” or some other word that provides little information about how it actually was. The person might continue with a more detailed description, but the initial word is usually to ambiguous to have real meaning.
Now imagine the person had instead responded by saying “my overall experience was 1.2 standard deviations above the mean,” or “I’d rank it in the 87th percentile.” You would have a lot more information about how the person felt about their experience. We could have the same type of improvements in things like nutrition facts (“.3 SDs more fat…”) or prices (.5 SDs less than competitors…).
To get back to Carey’s point, I think you could hypothetically teach things like distributions, percentiles, and z-scores to younger kids — they are straightforward enough concepts that middle or high school students could grasp them without needing to know the complicated algebra and calculus behind them. The problem is that it’s hard for the shift to come from within the school system. Schools tend to reflect and reinforce the societies they’re in. Statistics education won’t take off in schools until knowing statistics becomes a part of the random conversations and interactions in everyday life.
I do think it’s undeniable that at some point in the future middle school students will be learning advanced statistics. In their work on the “restructurations” that can take place in the way we learn, Seymour Papert and Uri Wilensky often discuss the historical example of the switch from Roman numerals to our current Hindu-Arabic number system. You’ve probably never tried it, but multiplication and division are difficult to do with Roman numerals. In those days there were not many people who could perform those operations, and that meant shopkeepers and merchants were generally unable to do their books themselves. Overall, it was a huge burden on society. Over the course of a few centuries, the adoption of the Hindu-Arabic system made it possible for nearly everybody to do arithmetic, something that was unthinkable for most of human existence.
These drastic shifts happen all the time. A few hundred years ago there were only two people in the world who knew how to do calculus. Today there are quite a few more. The role of statistics in an increasingly data-filled world will continue to grow, and one day people will look back on our current level of statistical proficiency as a relative dark age.