Connect with us
Click me!


Spinning the stats

Sans mathematical assistance, raw data can be misinterpreted. Unfortunately, for these two, the data had already been processed into a chart.

Dean Dela Paz



You would think that the two were looking at the same Rorschach ink blot and as expected both would simultaneously see different demons depending on underlying thought disorders, inner fears and subliminal psychological makeup.

But no. They were not looking at formless disembodied splatters. They were looking at simple bar charts derived from quantitative data untouched by statistical analysis.

One saw one wave. The other, two.

Abbott and Costello could not have written a more hilarious skit.

Both men were describing the exact same charts tracking the number of COVID-19 cases and deaths provided by the same technical working group of the Inter-Agency Task Force to which both belong. The charts are the same. The numbers, the same. Everything, every iota of data and detail, were exactly the same.

And yet, between these ranking Cabinet officials, one a Georgetown and Harvard-educated and pedigreed physician, the other, a former soldier, both cannot tell whether they were seeing one wave or two waves.

Raw data, unfiltered and unanalyzed through the usual mathematical and quantitative processes are like that. Sans mathematical assistance, raw data can be misinterpreted. Unfortunately, for these two, the data had already been processed into a chart.

What more when numbers are arrayed as statistics where critical analysis is needed from statistical operations spanning from averages, weighted averages, means, medians, modes and standard deviations, all simultaneously applied?

One can almost hear veins exploding in their heads.

What happens when high ranking presidential advisers cannot tell the difference between two digits enough to determine three percent margins of error? Would not everything effectively be in error?

Recently, a third ranking official stepped into the fray to settle the debate. He’s a lawyer.

Now it is our turn to succumb to panic. It’s like having the Village People (the band) take a Rorschach test.

It’s really then a matter of credibility. A question of who to believe. Note a recent fiasco involving data on American prime time TV.

In these times of quarantines and lockdowns where our principal accesses to the world outside are through social media, digital chat rooms and news networks on piped-in cable TV, for many, perhaps the only source beyond our shores comes from a poorly rating news network operating out of Atlanta, Georgia.

As in the case of Abbott and Costello, perspectives can unfortunately be skewed where data sources are severely limited. They not only skew perception; they seriously screw it.

Imagine viewing a vast ocean through a single porthole while trapped below deck sailing the high seas in a doomed ship infested with COVID-19.

This cable network, in a pubescent parody of itself, took virtual selfies and self-indulged in an ad it created. It tickled and fantasized itself a paragon of facts. Perhaps there’s some Freudian self-gratification in that. Never mind that theirs is farthest from reality.

The following are verbatim quotes on a network that spun Gallup data to claim two-thirds of Americans need a vaccine as a requisite to lifting lockdowns.

Reports show the network “misrepresented a Gallup poll’s results in an effort to bolster (a) narrative that most Americans would support a potentially years-long national lockdown.” The network claimed 68 percent of Americans said so.

Unfortunately, it was quickly debunked by the very data they referenced. Tagged an “egregious misreading of the data,” the claim failed to link which Gallup survey they were referring.

In a different Gallup poll results showed the opposite where comparable responses were only nine percent. Gallop had never said two-thirds of Americans needed a vaccine to return to work. Like the difference between one wave from two, 68 percent is not nine percent.

One problem with statistical charts is that the data are derived from sampling. Whether random or focused, these are technically not hard facts that can stand a 100 percent revalidation of each and every instance. They don’t. They remain interpretations. Politicians, soldiers, even lawyers should know better.

Click to comment