USA: esiste una relazione tra COVID-19 ed implementazione del 5G ?

vi rimando ad uno studio condotto da Magda Havas che ha fatto una analisi statistica tra
https://magdahavas.com/is-there-an-association-between-covid-19-cases-deaths-and-5g-in-the-united-states/

I eventually found which U.S. cities had 5G coverage as of April 21, 2020 and compared those locations to covid-19 statistics for April 22, 2020.

Commenti da MAGDA HAVAS 

Researchers in life sciences tend to use a probability (p) of 0.05 as a cut-off point (and sometimes 0.01).  If the value for the test is below  0.05 it means that the data are significantly different with a confidence of 95%.  The lower the p value the more confidence we have in the data.  For values above p >0.05 we state they are not statistically significant.

Consequently … 

1.  for the number of tests conducted for Covid-19 the p value was 0.4 which means that the data were NOT signficicantly different.  There was no difference in the average number of tests done per million population.  This is good news.  Had there been a difference here we would not be able to compare the data sets.

2.  For the Cases/million we had a p value of 0.02, which means the data are significantly different, i.e. more cases/million of Covid-19 in states with 5G.

3.  For the Deaths/million we had a p value of 0.05, which is also significantly different, i.e. more deaths/million for states with 5G compared to states without 5G.

So the result show that people in states with 5G have a greater risk of getting Covid-19 and dying from it.

This needs to be verified and other places need to be examined.  If this is indeed the case … we are in deep doodoo (not a scientific term)!


------------------------------------------------------------------------------------------------------------------------

altra indagine di ANGELA TSIANG (USA)

(Data sources: 

Cases, Deaths, Tests by State: https://www.worldometers.info/coronavirus/country/us/

mmWave states identified from https://www.androidcentral.com/heres-every-us-city-5g-coverage-right-now

Population density data: https://simple.wikipedia.org/wiki/List_of_U.S._states_by_population_density  

Air Quality Index:  http://www.usa.com/rank/us--air-quality-index--state-rank.htm )

LINK  al database di posta 

-------------------------------------------------------------------------------------------------------------------------------

8/5/2020

Update: I've been thinking about how to check if mmWave exposure is significantly correlated to the case and death rates and how to separate that from population density.  So I came up with a "mmWave exposure factor,"  calculated by adding up the populations of the cities with mmWaves in each state and dividing by the total population for that state.  Then I also checked for interaction of population density with this mmWave exposure factor, and it turns out this interaction is strongly correlated to the cases and death rates; i.e. population density acts synergistically with mmWave exposure in determining the cases and deaths rates.   Correlation of cases/million to this interaction is 0.703, and correlation of deaths/million to this interaction is 0.785.  See table below.

Population density is not only a measure of person to person contact, but also the amount of RF radiation a person is exposed to; for example, in higher population density areas, one is also exposed to higher levels of RF from their neighbor's routers, hot spots, and wireless devices.  Because RF exposure is inherent in population density, I sought a way to determine a mmWave exposure factor independent of population density.

Regression Models

Then I found regression models for cases and deaths rates.

When I ran regressions to model for cases/mil, deaths/mil, %cases/test, and %deaths/test by  including just population density as the one variable, they were not very good (R sq. very low).  When I ran regressions with just mmWave exposure, they were not very good either.  The regressions were best when they had population density as one variable, and either one of mmWave exposure or interaction between mmWave exposure and population density as the other variable.  The result was 4 decent regression models, with statistically significant p-values for all 3 variables (pop density, mmWave exposure, mmWave*population density) between 0.000 to 0.003.  The R sq. adj. was between 49% to 73%, and the R sq. predicted was between 22% to 47%

These models that incorporate mmwave exposure and the interaction between mmWave and population density were better than the ones calculated  based on population density and public transportation commute ratios.  In New York a consistent positive correlation between cases per capita and density  was calculated to be R-squared = 0.17, p < .01,  and cases per capita and public transportation commute ratios  was calculated to be R-squared = 0.25, p < .01).    https://medium.com/@cadre/does-density-determine-covid-19-destiny-lets-look-at-the-data-a43b60eac787  

In conclusion, I would say that the strong correlation factors and the regression modelling shows statistically significant evidence that mmWave exposure and the synergistic reaction between mmWave exposure and population density contribute to the increased cases/ million, deaths/million, %cases/test, and %deaths/test seen in mmWave states relative to the non-mmWave states.

Do we have any statisticians in the group?  If so, I would love to talk to you.

Angela Tsiang

USA