The largest value in Corona-times have to Pay – in addition to pasta and toilet paper – probably. People who have mathematics a little, and with statistics at all, nothing on the hat to call several times a day data on Coronavirus cases. Suddenly everyone seems to know what exponential growth is. And it felt like half of Germany, the thumb presses in Isolation, the curve – THE curve – may be quickly and sufficiently flatten. However, statisticians warn against relying too much on the Numbers.
Money plant
Corona and the real estate market: wait and see? Or just buy now?
The Figures available contain too few information, explains Katharina Schüller, founder of the Munich-based company Stat-Up and head of the working group “Statistical Literacy” of the German Statistical society. “They represent only a small part of reality, namely, the seriously Ill, a part of the easier for Patients with symptoms and a very small portion of people without signs of the disease, which were tested because they were suspected cases.”
If many others are infected or not, “we don’t know and can’t guess, only more or less justified,” writes Schüller in a contribution to the University forum digitization: “We know that each must be our model calculations wrong.” Nevertheless, the conclusions could be right.
The Robert Koch Institute (RKI) is the Supreme authority for infectious diseases in Germany, to get Numbers with Sars-CoV-2 infected people and dead, the the disease Covid-19, had, by the country authorities. In turn, you will receive from the regional health offices. “Between the notification by Doctors and laboratories to the health Department and the Transfer of the cases to the competent state authorities and the RKI for a few days may elapse (Reporting and transmission delay)”, says the RKI to. Numbers are also added later in the statistics.
Data “with caution” to use
Others, such as the Johns Hopkins University from the USA also include sources such as regional media. The European centre for disease prevention and control (ECDC), an Agency of the EU, compares, in turn, according to its own display state data with about 500 other sources, including Twitter, Facebook or Youtube accounts by ministries of health. “Users are advised to use all data with caution and taking into account their limitations of” extra on the ECDC website.
And also, the world health organization (WHO) indicates, among other things, on “differences in reporting methods, retrospective data consolidation and delays in financial reporting”. Because of the incubation period, the time for the Test and the reporting delays, for example, measures such as no-Contact orders show in Germany is often only about 14 days later to follow up in the Numbers.
There are many such stumbling blocks in the Corona data. The devil is, as so often in the Detail. “The number of the tested Infected not has to do only partly with the number of actual Infected, because people will be tested with few or no symptoms so far, in the rarest of cases, especially if you are Infected had no contact to be proven”, explain the creators of the “same UN statistics, of the month”, a range of several statistics experts, indicating a possible error in the Interpretation of statistics.
Sensitive-Country Comparisons
New test procedures could change and also, for example, previously unrecognized cases (keyword: dark baton) charge, but then the reported case, could increase numbers, without that, there is an accelerated disease dynamics.
A particularly tricky cross-country comparisons. “In particular, the collected case depend on numbers in each country of Central of how systematically and extensively the Virus is tested,” explain the creators of the “same UN statistics,” which Katharina Schüller.
A number of factors influence the level and severity of the infection and can vary from country to country is immense: to test number of inhabitants, age structure, specific diseases in the population, such as tuberculosis, the stage of the outbreak of the wave, the will or the assets of, the guidelines for who is ever tested. In old homes of deceased people not to be tested in some countries afterwards, and are included in the statistics – in other. Because it is mainly Older die with Covid-19, can have the huge apparent differences.
The statistical detection of the causes of death vary greatly from country to country, underline the authors of the “same UN statistics”. However, comparisons of mortality rates are discussed. Generally speaking, it’s wrong to put the dead in relation to the known Infected. The dark figure did not take into account (which in turn depends significantly on the extent of the Tests performed), then the denominator is going to – so the number under the fraction bar when you Share – it is too small. It follows that The estimated case fatality rate – i.e. the proportion of deaths among all Infected is systematically overestimated.
The exact Figures to create the illusion of accuracy
Tricky, it is also used in statements to the number of Recovering, here and there given up to the last spot and so pretty exactly the appearance. But where not even all of the Infected and collected to be tested, of course, can be much less than the number of Recovering known. Therefore, all of these figures are only estimates – and very rough estimates in many cases.
The Bavarian state office for health and food safety, explains: “cases are considered to be recover, if the reporting date no longer than two weeks and no hospitalization is present without a known release date, no pneumonia and/or dyspnoea, and no case of death was reported to be present.” Here are some launched surface to help deckendere Tests, among others, to antibodies, to improve the estimates and projections, at least a little.