Jul 1, 2014

XL-Files: The Future of Statistics…?

Xiao-Li Meng writes:
Life sometimes teases us. The older one gets, and hence the shorter future one has to contemplate, the more often one is asked to predict the future. I keep being invited to speculate about our future. But asking a statistician to predict the future of statistics is as cruel as asking a barber to cut his own hair or a surgeon to operate on herself, or worse, asking sausage makers to consume their own sausages!

Fortunately, as statisticians we can proudly say “I’m not sure,” and “Let me collect some data!” Recently, I’ve had opportunities for discussions with several groups about our future in academia. Five possible states emerged, as synthesized below. These “wisdoms of crowds” came with neither citation nor validation. Readers therefore must apply their favorite methods to separate noise from signal, if any, and to determine which was/is the modal state (leave your comment below!).

So should/would we…

Fear OR’s Minimization?
The prominence of Operational or Operations Research (OR) grew during WWII, peaked around the 1960–70s, and declined steadily since then, at least in the popular media. Many pointed out that what have been minimized are not the methods, techniques, or even the theories developed in OR—they are as popular as ever and encompass most of the optimization methods we routinely use, for example. Rather, it is the brand name, OR, that has been overshadowed by sexier labels such as Management Science or Business Analytics. But isn’t this exactly the worry some of us share that the brand name Statistics, not its substance, will be overshadowed by sexier labels such as Data Science or Predictive Analytics?

Establish Mathematics’ Justification?
A more optimistic opinion is that statistics will enter a state that mathematics has enjoyed for a long time (and likely will indefinitely). Although our society often portrays mathematics as esoteric, few would suggest eliminating mathematics departments even in hard times; this threat remains real for some statistics departments (particularly in some countries), though much less so now than just a decade ago. What did/do mathematicians do to establish and justify that they are indispensable? No learned person would dismiss the centrality of mathematics in the evolution of science and civilization. Nor would any responsible parent dissuade their children from acquiring basic mathematical training even if they themselves flunked it. Are we about to enter an era where the same statement continues to apply when we replace “mathematics” by “statistics”?

Achieve Biology’s Diversification?
An even more ambitious prediction came from witnessing how the field of biology has evolved since the ’50s into a full Division or School of Life/Biological Sciences. For example, at the two universities where I have taught, Chicago has at least five departments involving the B-word (Microbiology, Neuro-biology, BMB, MGCB, OBA), and Harvard has at least six (Systems Biology, CCB, HEB, MCB, OEB, SCRB), not counting many more such as “Ecology and Evolution” and “Human Genetics”. Could there be a future for a Division or School of Statistical/Data Sciences, hosting as many departments but with the S-word: Astrostatistics, Geostatistics, Engineering Statistics, Financial Statistics, etc., without even counting various -informatics or -metrics?

Follow MBA’s Professionalization?
Of course as far as degrees go, nothing is more popular or populated in our society than the professional degrees such as MBA. Statistics has now grabbed our society’s attention, or at least that of high school and undergraduate students, to a degree far exceeding, almost surely, any statistician’s dreams just half a decade ago. For example, about 20% of Harvard Freshmen this year listed statistics as a potential concentration/major choice and we already have 160 concentrators in-house, not counting any secondary fields/minors. If I opened a bottle of champagne for every 2% increase—the bar I set for myself six years ago—I’d be too intoxicated to finish this sentence. Could this dramatically increased attention eventually transform into something truly intoxicating (with the buzz and hangover both included)?
Employer: “Sorry, but you don’t have an MDA.”
Applicant: “I do have an MBA, it’s on my CV.”
Employer: “No, you don’t. I mean MDA, Master of Data Analysis.”

Enjoy Physics’ Unification?
Compared with biology, physics has more presence in our universe(s) but many fewer departments in our universities, even though it requires as much lab work. For example, Chicago only has two departments with “physics” in their titles (Physics, Astronomy and Astrophysics) and Harvard has one department (Physics) and one degree committee (Biophysics), a situation not unlike having Statistics and Biostatistics. Could it be that physicists’ mentality of finding a unified theory of everything unites their field, from experimental physicists to string theorists? If so, can our desire to find a unification of Bayesian, Frequentist and Fiducial (BFF) perspectives do the same trick, allowing all of us to thrive under one roof as BFFs (Best Friends Forever)?

1 Comment

  • The rise of statistics in the society has certainly been a phenomenon coupled with the surge of Big Data in its popularity. The trend is a good thing to the extent that statistics and data literacy are reaching a wider audience than it used to be. Many start realizing the importance of using data and digging “Gold” in the new “Data Rush”: businesses and governments look for insights for making data-driven decisions, and invest a plethora of money to build infrastructure and hire talents; fervent career-changers and students hope to look in “sexier” shape when searching for jobs and start taking courses; and schools respond with more and more educational resources in teaching the subject.

    But I am skeptical about the “Big Data” scene in that much of the interest, however, seems to be in the use of statistics, not its substance. By substance I mean the theories and methods. I observe that outside academia, there is a lot of preference of “whatever works” over “a correctly applied model or method”. From my analyst experience, this pragmatism works when it happens to work but can also throw me into confusion when it does not. There are definitely two sides of the coin.

    While I don’t deny that Data Science truly has something new to offer, I feel that there may be too many gleaming words like Data Science, Predictive Analytics, etc, around conferences and events. I even heard of some speaker at a conference advocating that data science is revolutionary in making us contemplative about ourselves as human. Nevertheless few of these mention statistics, to my surprise. But in fact, a good part of the new glory is not new but rather usual “business” regarded in statistics, OR, or computer science. Would someone make the same bold statement about data years ago?

    I like to believe that after some points, the hot trend of Big Data will gradually slow down and then return to what it’s worth. But right now, the trend will be climbing up and attracting a lot of interests for a while. And I think statistics definitely has a large role to play in harnessing these interests and educating the wider audience.

Leave a comment



Welcome to the IMS Bulletin website! We are developing the way we communicate news and information more effectively with members. The print Bulletin is still with us (free with IMS membership), and still available as a PDF to download, but in addition, we are placing some of the news, columns and articles on this blog site, which will allow you the opportunity to interact more. We are always keen to hear from IMS members, and encourage you to write articles and reports that other IMS members would find interesting. Contact the IMS Bulletin at bulletin@imstat.org

What is “Open Forum”?

In the Open Forum, any IMS member can propose a topic for discussion. Email your subject and an opening paragraph (to bulletin@imstat.org) and we'll post it to start off the discussion. Other readers can join in the debate by commenting on the post. Search other Open Forum posts by using the Open Forum category link below. Start a discussion today!

About IMS

The Institute of Mathematical Statistics is an international scholarly society devoted to the development and dissemination of the theory and applications of statistics and probability. We have about 4,500 members around the world. Visit IMS at http://imstat.org
Latest Issue