Hi Xin Wei,
The first day in engineering school, the statics and mechanics professor told us a little joke: during the activities around commencement, they set up a philosopher, a physicist and an engineer for a 100 m race. When the gun went off, the engineer sped off while the philosopher and the physicist stayed back to argue Zeno’s paradox as to whether or not they would ever get to the finish line. When the engineer sauntered back to the start line with his prize, the philosopher and the physicist looked up in surprise, the engineer shrugged his shoulders and said “Close enough is good enough for me”. I’m not a statistician or an epidemiologist but I try to look at numbers like I read. In the same way that I can read your short test in this posting as the tip of an iceberg, I can go on to problematize your statements, complexify the ideas, compare your words to others of the same ilk. On their own, they’re just another opinion. And it is only through consideration of what is at play on various levels with other writings, traditions, thinkers that you short paragraph gains worth.
I’m no train watcher, but I have been following the “official” numbers since the end of February out of frustration with the poverty of analysis which popular media and unpopular media give us and also out of a healthy mistrust for the official story. I have been tabulating and comparing side-by-side the numbers for 8 territories of interest to me and what the numbers tell me always goes beyond the official analysis and popular thought. Even with imperfect numbers, one cannot give our decision makers the exclusive domain on their interpretation. Further, comparing statistics, flawed or otherwise, have helped me navigate through what is often political propaganda and self-serving deliberate misinformation posing as wisdom and enlightened insight. I hardly call this a useless exercise.
I know that in culling these numbers, I am dealing with an incomplete and imperfect data set and that they are open to manipulation. Even if they are the tip of the iceberg, I know that these numbers are like best case scenarios, in that you can’t subtract from confirmed cases and that there are likely a lot more cases than what we are being told. And if that’s the case, then the first conclusion to arrive at is that the problem is bigger than we are being led to believe, and that we ought to be even more concerned. Still, these imperfect or limited numbers lead me to ponder about possible causes or explanations about why things are happening as they are. One has to act on something, and alter course as necessary. Nate Silver’s analysis, hedges its bets at every opportunity, and other than encourage us to question the accuracy of the numbers, doesn’t come up with any cogent strategy as to how to make the existing numbers more reliable, or more trustworthy. I like that it instills a healthy skepticism for their origin, yet it is coming rather late in the game… but better late than never.
I don’t know about you, but for me, 350,000 confirmed cases is not meaningless. Is it really 400? 500? In five days time will that be 600,000, 700,000 a million? For me examining and comparing the numbers is a machine for thought-creation or question-posing. Why are there four times the number of cases in the US than in China so far? Why are only China, Russia and Iran numbers questionable? Or why is there a disparity in the number of deaths per thousand cases between the various territories? Or what leads to spikes in the number of new cases? Or why is there a knuckle in the log curve of new cases versus date in the US or other countries? Or why there are concentrations of contagion in the peripheral communities of NYC? Or why the data might be disproving a politician’s assertions about the pandemic? The numbers don’t tell you the story--they are not the pragmatic outcome. But they do condition and enrich any upcoming encounters with new information in terms of being able to problematize it and to establish connections through the complexification of thought about the pandemic. My conclusions might not be always correct, but at least they allow a deeper understanding through the emendation of preliminary opinions, interpretations, or deductions. For me the numbers are not “reality” but stand-ins, signs of what possible movements might be happening behind the pictorial surface of bare cases. No more, no less. As Whitehead might say, they are lures for feeling that develop with the concrescent phases of the subject in question.
China called for lockdown in 01.23 when they had 260 new confirmed cases per day. At the same time, Trump already knew there was a case in the US, “It’s one person coming from China, and we have it under control”. The USA was still prevaricating with lockdown when they had 25,000 new confirmed cases per day, almost two and half months later. At this point, do I really care whether the degree of confidence in the numbers is in the oder of +/- 20% because of flawed testing? Or that the numbers may be imperfect because different testing strategies can impact the number of reported cases? Do I get pissed off to hear that Legault in Quebec is encouraging the general population to continue going to work on the week of 03.16, with 200 confirmed cases and a string of day-to-day percentage increases of 20%? Do I scream in disbelief on 03.08 when the Government of Spain calls on women to go out and march in solidarity, with 430 confirmed cases after a string of day-to-day increases of over 30%? I don’t need an epidemiologist to explain to me that the numbers may be flawed in order for me to become an indignado.
One doesn’t need a PhD in mathematics or statistical analysis to get that if the weather forecast calls for a 60% probability of rain showers for your city that day, it makes sense for you to close the windows of you home before you leave for work. If we know that the actual numbers are greater than the detected or confirmed, and that the pandemic is making its way westward, as soon as it is detected, one knows the pandemic is established and that the progression will have a life of its own, unless decisive action is taken. If the object of the game is to get the basic reproduction ratio R to below 1, and lockdown is the most effective way to stop the spread, why didn't they call for a lockdown as soon as the first case appears, if we know and understand the dynamics of contagion and its progression? Flatten the curve for how long? How much faith will people have in the tail-end of a flattened curve to go out into the world and risk contagion? The Pandemic of 1918 lasted into 1920... To me current strategies sounds too much like Russian roulette for expendables… People might say that I am a polemicist, or activist, but at least I'm not serving as an apologist for what's happening. And one thing I'm not doing is Monday morning quarterbacking, I've been calling it as it was happening.
I think you sell your audience short as to their abilities to read numbers and over-estimate their degree of trust in statistics. If there is something that is saving the US and Brazil it is that the intelligence of the social is prevailing over the stupidity, miscreancy and self-interest of our so-called leaders, decision-makers, and media pundits.
--Technological solutions such as machine learning statistical modeling are limited by the quality of the data in which they are based (a scientific methodological point typically overlooked by programmers and armchair epidemiologists :):) ). and in this case the tables of coronavirus statistics are practically useless without understanding how they were derived.
Numerology : Mistaking a numeral — a squiggle on a screen or a piece of paper — for a fact of the world.
___________________________________________Sha Xin Wei | skype: shaxinwei | mobile: +1-650-815-9962 | asu.zoom.us/my/shaxinwei________________________________________________APR. 4, 2020, AT 1:11 PM
Coronavirus Case Counts Are Meaningless*
*Unless you know something about testing. And even then, it gets complicated.
By Nate Silver
If you’re a regular reader of FiveThirtyEight, you’re probably used to looking at data in sports — where basically everything that happens on a basketball court or a baseball diamond is recorded — or in electoral politics, when polls (in theory, anyway) survey a random sample of the population. COVID-19 statistics, especially the number of reported cases, are not at all like that. The data, at best, is highly incomplete, and often the tip of the iceberg for much larger problems. And data on tests and the number of reported cases is highly nonrandom. In many parts of the world today, health authorities are still trying to triage the situation with a limited number of tests available. Their goal in testing is often to allocate scarce medical care to the patients who most need it — rather than to create a comprehensive dataset for epidemiologists and statisticians to study.
Research Stream: http://synthesis@posterous.com
Home: http://synthesis.ame.asu.edu