Lies, Damned Lies and Statistics

I was looking at the Local news this morning and found this article.
Gov. Wolf Says Local Counties With Percent Positivity Above 5 Percent are 'Cause for Concern'
pittsburgh.cbslocal.com...

I'm thinking to myself 5% that doesn't sound right. So I went to the State Covid-19 Dashboard site.
www.health.pa.gov...

The article mentions six Counties. Beaver, Allegheny, Washington, Indiana, Lawrence and Westmoreland. I started doing a little math. I took the number of cases per county from the Dashboard and divided it by the population. The most recent population numbers that I could find are from the State Data Center here. pasdc.hbg.psu.edu... I looked a bit more and found that the "Cases as of 7/17/2020" number was a cumulative number since March and that the State was estimating a 76% recovery rate. This dropped percentage even lower. The "Active Cases 7/17/2020" was determined by subtracting the 76% cured from the "Cases as of 7/17/2020". The % After Recovery" was determined by dividing "Active Cases 7/17/2020" by the Population.

What I found was:
County Cases as of 7/17/2020 Population Total % Active cases 7/17/2020 % After Recovery

Beaver 949 164,742 0.058% 323 0.019%

Allegheny 6,125 1,218,452 0.050% 2,083 0.017%

Washington 540 207,346 0.026% 184 0.0088%

Indiana 97 84,501 0.011% 33 0.0039%

Lawrence 225 86,184 0.026% 77 0.0089%

Westmoreland 1,135 350,611 0.032% 386 0.0011%

There's some real small percentages there. Nothing anywhere near 5 percent. I was figuring that I was on to something and was starting a post on ATS until I re-read the article and found out that I had made a mistake. The percentages being referred to were not the percent of positive cases. I had used the wrong Dashboard. The one in the article was the "Early Warning Monitoring System Dashboard".
www.health.pa.gov...

It had a different set of metrics. OK my fault for not reading closer and jumping to conclusions. This Dashboard reported on things like "Difference in confirmed cases (most recent 7 days vs previous 7 days)", Incidence rate (most recent 7 days) per 100,000 residents", PCR percent positivity (most recent 7 days), difference in the average daily number of COVID-19 hospitalizations, Difference in the average daily number of COVID-19 patients on ventilators and Percentage of hospital emergency department visits in the most recent 7 days due to COVID-like-illness. I was getting ready to scrap my post until I remembered something from a Statistics class I had a long time ago. The instructor was talking about polls. He made the statement that is the title of this post. Lies, damned lies and statistics. I realized that my original numbers made a case for showing how blown out of proportion the State's response is. They realized it too. That's why they made a new system of measurements to justify their agenda. The first three measurements are related to testing. With the increase in testing those numbers would normally show an increase if negative tests were not factored in. The next two are determined by medical coding, the parameters of which are determined by the State. (there's where your regular flu cases went) and the last has it all in the title "due to COVID-like-illness" Have you seen the list of symptoms for COVID-19? I could break my arm and qualify in that category.

Another thing to notice is that all of these statistics are based on a period of 7 days or a daily basis. There is no way to show trends over a period of time.

We are being scammed. Plain and simple.

a reply to: JIMC5499

State-istiscs are used to control the pop.

a reply to: muzzleflash

Read the rest of it. I accidentally posted before I was done.

Too many people in this world don't know how to statisculate. Sadly, even more don't seem to be able and/or willing to speak the truth.

Statistics have always been manipulated and spun. But like scientific studies, numbers don't lie, in fact numbers are absolutely true and factual. One plus one will always equal three.

Now repeat after your respected authority, "One plus one equals three." Let's hear that again, but louder this time.

a reply to: JIMC5499

I went and got drunk last night, and the mini hangover I had this morning is all because of Covid-19.

a reply to: AutomateThis1

You could probably make a case for it. Decompressing from all of the bull# stress inflicted by the overreaction to the virus.

The statistics are wrong. There are many more incidences of covid19 that just haven't been reported.

When school starts there will be a second wave that will give a new meaning to pandemic. Masks, hand sanitizer. and 6 ft. distancing won't do jack sh!t.

originally posted by: muzzleflash
a reply to: JIMC5499

State-istiscs are used to control the pop.

What's astonishing is how QUICKLY populations across the globe turned to jelly and tugged their forelocks. THAT'S far more disturbing than anything Covid-19 itself has thrown at us.

A fact that future potential demagogues and shysters will remember with chin-stroking gusto.

The quote attributed to Hitler that world leaders should be aware that the populace "does not think", although never proved to be his own words, nevertheless still seems pretty accurate 80-90 years later.

originally posted by: muzzleflash
a reply to: JIMC5499

State-istiscs are used to control the pop.

Yes, to mesmerize the gullible.

Your calculations are off by an order of magnitude.
In your first example, Beaver, the correct cases% is 0.58% (not 0.058%), active % is 0.19% (not 0.019%)
In your last example the correct percentages are 0.32% (not 0.032%) and 0.11% (not 0.0011%)
NUMBERS DON'T LIE --> 0.011 is 1.1% - 0.0011 is 0.11%

ganjoa

a reply to: ganjoa

I stand corrected. It's still a small number.