Too early to tell

Re: “Who believes the stats?” (PostBag, March 29).

Khun Burin Kantabutra in his observations on the death rate related to the Covid-19 pandemic (Bangkok Post, March 29, 2020) seems puzzled by the low mortality rate for Thailand (0.38% by his calculation). Perhaps he should notice that it is not that unusual — several countries listed in the tabulation provided in the March 29 Bangkok Post have similar rates. For instance, Austria 0.7%, Norway 0.5%, Czech Republic 0.3%, Chile 0.3%, Israel 0.3%, Germany 0.7%, etc. What is significant are the low figures for Germany and Austria compared to those for e.g. Italy. Both countries report a large number of deaths (in the thousands, whereas other countries listed report very small numbers). Austria and Germany are neighbours of Italy, and have close cultural links with that country (Lombardy, the most affected part of Italy, was until the 19th century part of the Austrian Empire). Why Austria and Germany report much lower mortality rates than Italy is not known and is subject of an interesting article in The New York Times on the same day.

On the other hand, the low figures reported for other countries (including Thailand) are most likely statistical flukes. Even a relatively small increase in reported deaths could significantly change these figures. Complete analysis of data collected during the pandemic will certainly reveal some interesting patterns but now it is too early to draw firm conclusions. Similarly, it is very perilous to make predictions based on small samples; such predictions have an inevitably low level of confidence (using statistical jargon).

PRZEMO KRANZ