Know what? It doesn't...
SURE it can be and is meaningful - IF a random selection are 'culled' from the group - or a process excludes a certain random (BASED on no criteria which otherwise skew the result) part of the population from consideration in the first place.Well, there sure as heck appears to be not much in the way of randomness per that footnote at the bottom. If it was truly a "random" cull then the entire population would be run through some type of bootstrap resampling method to insure the degree of randomness.
It's called 'sampling' and it's done all the time in a multitude of fields from semiconductors testing to medicine ...Yea, and "sampling" is done to massage results or come up with desired plots. It's a weak histogram. A sampling scheme with an iota of respectability would strive to retain the largest viable block of the population in question, not the exact opposite.
Also, the "sampling" you cite for medicine or semiconductors QA\QC is totally different from a statistical analysis designed to project a rate of increase in infection and mortality based upon the onset, population and death toll of an infectious disease.
Take a look at the top 6 countries. Do you notice anything odd, or makes you go hmmmm???
Full Chart Below
.
.
.
* Includes number of deaths, recoveries and current cases
1. WHO accuses China of underreporting cases in Beijing
2. Includes suspected or probable cases
3. U.S.A. is only reporting suspected cases under investigation.
Numbers in red indicate an increase in the last 24 hours.
Numbers are not updated on Saturday or Sunday
Source: Reuters
or what?
I don't think you really have ANY basis for making a valid objection.
I'm sticking by what I wrote/what I posted.