Data analyst says Turkey’s Coronavirus figures are inaccurate

Turkey’s daily figures for new Coronavirus cases have recently started to be published again, but one data scientist has suggested that they continue to be inaccurate.

Nick Brown looked at the figures released by the ministry of health, and found that the total figures for cases, deaths, tests and recovered patients rarely end in 0 (zero), which is statistically unlikely to happen randomly.

Brown also said that the figures themselves were inconsistent in their formatting

He concluded that although it was uncertain what could have caused this anomaly in the figures,”it does seem somewhat unlikely that these numbers are entirely the product of natural processes. Perhaps some artefact of the data collection protocols is causing this.”

Brown invoked the statistical principle of Benford’s law, which looks at the frequency distribution of digits in real life data sets, to show that it was unlikely that the Turkish Health Ministry’s figures were arrived at naturally. One respondent to the thread noted that “Bedford's Law roughly held in China, US, and Italy”, linking to an academic paper which analysed the figures from those countries.

On Friday Turkey posted its highest daily infection rate numbers, with almost 30,000 new cases. Doctors and other medical practitioners have been warning for months that the number of cases of Covid-19 in Turkey is far worse than reported.