Read on Twitter
THREAD: (1) Many want to use country comparative Covid19 case/death statistics to rate/compare Covid19 policy in different countries. This is complicated, as it is uncertain how policies affect these statistics, and as it is uncertain how comparative the statistics are. =>
=> (2) There are so many factors besides pandemic policy determining how the pandemic spreads and kills, and we are so far from the end of this sad story to know the sum of it all, but I understand that people seek a sense of safety, a feeling of control, in this way. I do too.=>
=> (3) So here's a few tips to make the comparison you do a bit less random and slightly more meaningful:
(a) Use Covid19 charts (cases/deaths) that adjust for time, as different countries are in different stages of the pandemic: such as this one: https://ourworldindata.org/grapher/covid-daily-deaths-trajectory-per-million =>
(a) Use Covid19 charts (cases/deaths) that adjust for time, as different countries are in different stages of the pandemic: such as this one: https://ourworldindata.org/grapher/covid-daily-deaths-trajectory-per-million =>
=> (4)
(b) Use charts that adjust for population size (eg. deaths per million), as absolute numbers for countries with very different numbers of citizens will be comparatively misleading.
(c) Use charts with rolling averages to adjust for random flaws in daily reporting. b+c =>
(b) Use charts that adjust for population size (eg. deaths per million), as absolute numbers for countries with very different numbers of citizens will be comparatively misleading.
(c) Use charts with rolling averages to adjust for random flaws in daily reporting. b+c =>
(5) An example combining the properties a-c is this one: https://ourworldindata.org/grapher/covid-daily-deaths-trajectory-per-million
(d) Use logaritmic curves rather than linear, as all countries have similar shapes of the linear curve of increasing numbers, but different rates of this increase, and the rate of =>
(d) Use logaritmic curves rather than linear, as all countries have similar shapes of the linear curve of increasing numbers, but different rates of this increase, and the rate of =>
(6) => increase is a better guide to how the pandemic develops. Again, this is an example: https://ourworldindata.org/grapher/covid-daily-deaths-trajectory-per-million
BUT MOST IMPORTANT, YOU NEED TO MATCH THE COVID19 CURVES TO POLICIES, AND THE TIMES FOR THESE.
Pandemic statistics has to be combined with political statistics =>
BUT MOST IMPORTANT, YOU NEED TO MATCH THE COVID19 CURVES TO POLICIES, AND THE TIMES FOR THESE.
Pandemic statistics has to be combined with political statistics =>
=> (7) For instance, this source measures the stringency of the policy response in different countries in relation to time: https://covidtracker.bsg.ox.ac.uk/stringency-map
When you have to countries you want to compare based on the covid19 statistics, you should check their respective policy history =>
When you have to countries you want to compare based on the covid19 statistics, you should check their respective policy history =>
(8) => in order to see if ones with comparatively more favorable pandemic developments are systematically linked to a certain policy pattern compared to ones doing less well. If no such pattern can be found, we may conclude that so far no clear link between stringency of policy=>
(9) => and pandemic development can be based in available data.
END THREAD
END THREAD
