Pandemic teaches us the value of statistics

EVERY high-school mathematics teacher, and nearly every parent, is familiar with this complaint: “But why do I need to study algebra (or trigonometry, or calculus)? I’ll never use it again!” And we all know that the kid is probably right.

Unless he or she is destined to be an engineer or scientist, the use of high-level math probably will not be a requirement, at work or at home.

But the same is not true when it comes to statistics, a basic understanding of which should be required of every degree-seeking university student.

Statistical literacy is very useful in life, whether one is a dentist, carpenter, lawyer or street-sweeper. The covid-19 pandemic reminds us of this, in spades.

Every day, we are inundated with statistics about the number of cases and deaths caused by Covid-19. Medical experts come on television and display fatality percentages, flattening of curves, and rates of recovery.

But in order to fully understand the significance of all this, one needs to understand statistics, and statistical reasoning. For instance, we see in the news that India is experiencing the very worst of the pandemic spread – measured in new cases, deaths, lack of oxygen and slowness in vaccinations.

But understanding the extent of the horrors suffered there, when compared with suffering elsewhere, we need to think statistically.

News shows and newspapers usually do the work for us, but we need to do more than hear the numbers – we need to understand them, and this requires some level of statistical literacy.

Likewise, in order to reach sound decisions in everyday life – whether one is buying life insurance, talking with a stock-broker, deciding on an old-age annuity plan, or simply playing the lottery – understanding statistics can be very helpful.

Consider a decision involving a defined-benefit retirement plan. Should you opt for a retirement policy which guarantees payments for the worker only, until death, versus one which allows one’s spouse to collect if the worker dies first?

The latter choice means lower payments during retirement, but which one would be the better choice?

It is basically a bet on how long the worker and his spouse will live. Needless to say, a statistical analysis involving life-span averages and an analysis of how long each of the two is likely to live, given each one’s health and family history along with a comparison of pay-outs in various scenarios, is possible – provided one knows something about statistics.

Even lottery players might make wiser decisions by relying on statistical calculations. For instance, consider the (illegal) “numbers game,” still played in some places, where players bet on the final three digits of the amount wagered at a racetrack on a given day.

There are 1,000 possible outcomes, from 000 to 999. The prize is US$600, so a player would get back only 60% of a payout with true odds. Simple enough.

Some legal numbers-based games of chance, however, are much more complicated.

For instance, many lottery games feature cumulative “jackpots,” where, if no bettor wins the jackpot in a given draw, the jackpot portion of the bets is applied to the following draw. If this continues – with no jackpot won over several draws – the odds in subsequent draws improves considerably.

At some point, it is perfectly possible that the payoff in a given draw could exceed the amount bet by all players for that draw, because of the money applied from earlier draws where no jackpot was won.

So a thoughtful player might sit out early draws and buy tickets only when the jackpot payout offers players a statistical advantage.

Or how about the purchase of an automobile? Suppose a car dealer offers an extended warranty. Most buyers mulling such an offer probably do a very general (and inexact) calculation in their heads about whether it is a deal worth taking.

But if one applies statistical reasoning, a rational decision is easy to reach. First, how much does the extension cost? Second, how much will it cost if something goes wrong, anything from a fuel-line problem to engine failure? And how likely are any of these problems with a car aged, say, five years? Google can supply that information in a matter of minutes.

A car buyer armed with an understanding of statistics could gather the pertinent information, do some quick calculations, and come up with a financially responsible decision in a few hours.

So the next time your son or daughter complains about the uselessness of an advanced math class, just say, “Well, you’re probably right. But you need to pass the course to get your degree. Besides, figuring out math problems is good mental exercise. Oh, and by the way, take a statistics course or two.”

William G . Borges is a professor in the American-Canadian program at HELP University in Malaysia. Comments: