By Basab Dasgupta
Even though I have a PhD degree in theoretical physics I never took a course in Statistics in high school nor in undergraduate college. I studied statistical mechanics and quantum mechanics in graduate school which exposed me to statistical principles and especially the concept of probability.
I liked the ideas of statistics and saw the relevance to phenomena in our daily life, but I must confess that I was never completely comfortable in solving statistical problems. I understood the logic and derivation but always had some doubt regarding which logic or theorem was best suited to solve a given problem. Now that I am tutoring math during my retired life this has become a handicap.
Unfortunately for me, I am seeing a slow but steady infusion of statistics not only in high school curricula in the US but even extending to middle school. In fact, in many high schools, statistics is offered as an alternative to calculus in satisfying the graduation requirement. I loved to tell all my math students, “you studied pre-algebra, geometry, algebra, trigonometry and pre-calculus to prepare yourself for your ultimate education in math – calculus.” That message now seems to be outdated.
It is frustrating because I am very passionate about it. I get enormous joy in explaining concepts of “limits”, “derivatives”, “Riemann sums” etc. to my students of calculus and even greater joy when they seem to get it and appreciate it. It is unthinkable for me that a student graduates from high school without experiencing this joy.
Perhaps statistics is a perfect subject for American students to learn because the entire country and its economy run on statistics. Take one of the most profitable businesses in the nation, the insurance business, as an example.
The business models of all insurance companies assume by analyzing existing data that the probability of the mishap (death, illness, car accident, property damage etc.) is relatively small. As a result, the insurance companies can make big profit by collecting premiums from everyone even after paying out large compensations for victims who are affected. If the actual disaster rates exceed the numbers assumed in the model for some reason, then the companies start losing money and some eventually leave the business. This happened in California when multiple forest fires forced home insurance companies out of business.
Same considerations apply in banking business. Banks make money by offering relatively low interests on savings accounts but charging much more in loan interests and other financial services. They target clienteles who save money and at the same time require those services. A statistical distribution of such clienteles allows them to decide the best locations for branches.
Any organization or company preparing a financial budget for an upcoming year or quarter must make assumptions about sales, production, pricing etc. These are not blind guesses but based on projections from statistical analyses.
Analysis of data in medical research is heavily dependent on statistics. FDA approval of new drugs is determined by the statistical success rates of the drugs in treating the targeted illness.
There is always talk about “age group”, whether one discusses health care or income which is a statistical group for processing data. Census is nothing but a statistical analysis of massive amounts of data. Investment in stocks of a company is dictated by the probability that the company would be profitable.
Grades like A, B, C in all academic institutions are based on a statistical “bell-shaped” curve. Use of statistics is extensive in professional sports. The performance of each player is tracked, and their numbers are statistically analyzed to determine the value of the player which is used during draft or trade. These numbers remain at the fingertip of the broadcasters who often recite them during commentary of a game.
“Profiling” a person by the FBI is basically a statistical analysis to correlate people’s characteristics and habits with crimes they commit. It is not necessarily valid for one criminal but does turn out to be valid when applied to a large group of similar people. Similar profiling is used by match.com and other similar websites to bring compatible people together.
It seems that some statistical concepts are ingrained among the general population. I remember trying to explain the meaning of probability to a middle-school student using an example of rolling dice in Las Vegas. I anticipated it to be a difficult topic to grasp for this young girl. “You see, dice have six faces with six different numbers on them. If you roll a dice, there is equal probability that any of these numbers will be on top; the chance that you would roll a “two” is one in six. In statistics, we say that the probability of rolling a two is 1/6.”
“Oh, you mean, the odds of rolling a two?” She instantly understood the concept without any problem.
I attributed my lack of passion for statistics with my Indian background; perhaps my faith in calculus comes from a steady unshakable focused “Indian” mindset as opposed to wishy-washy fluctuating and indecisive western culture.
After all, India is the country of Aryabhata, Ramanujam, Satyen Bose and S. Chandrashekhar, not to mention an army of IT engineers keeping the global communication running smoothly where results must be precise.
Well, I was wrong. I was both delighted and surprised at the recent news of the Indian American statistician Dr. Calyampudi Radhakrishna Rao being awarded the biannual “International Prize in Statistics” which is supposedly equivalent to the Nobel prize in science, and only the fourth person to receive this award. I feel embarrassed because I didn’t even know his name nor his contributions to statistics.
The American Statistical Institute calls him a “living legend” whose work has influenced fields as diverse as economics, genetics, anthropology, geology, national planning, demography, biometrics and medicine. Dr. Rao was awarded the “National Medal of Science” by President George W. Bush and “Padma Vibhushan” by the Indian government among numerous other awards, medals and honors.
In particular, he is best-known for the Cramer-Rao bound and the Rao-Blackwell theorem, both related to the quality of estimators as well as the Fisher-Rao theorem, Rao distance and orthogonal array.
I was happy to see that Indians are equally strong in statistics as they are in science and mathematics.
My weakness in statistics has nothing to do with my Indian origin but stems from a flaw in our college curriculum. I now believe that Statistics must be offered with the same importance as the other STEM subjects, starting in high school and certainly at college.
Basab Dasgupta has a doctorate in physics from University of Wisconsin and worked with Sony as Vice President of an operating division. Retired, he now lives in San Clemente, CA.
Disclaimer: The views expressed are not necessarily those of The South Asian Times