When Growth is Presented as Decline

Hetq.am reports: “Yesterday, Health Minister Arsen Torosyan posted a graph on his official Facebook page regarding the spread of coronavirus in Armenia, which clearly depicted a sharp decline trend.

However, in the context of the daily increase in confirmed coronavirus cases, the contradiction of this graph is evident at first glance. What calculations were made to arrive at such a significant decline trend? By comparing the numbers in Arsen Torosyan's published graph with the actual data of confirmed cases, it became clear that the following formula was used in Torosyan's graph: (day 2 - day 1) / day 1 X 100%. The use of this formula to present the resulting numbers as a decline trend does not only fail to reflect the reality, but also misleads the public by describing growth as a decline.

The news outlet notes that the percentage increase formula used by the minister is incorrect in this case, as we are dealing with very small and very large numbers. In other words, the percentage change between the numbers 1 and 4 is 300%, while the change between the numbers 700 and 800 is 14%. However, the actual picture shows that when the number of infected individuals rose from 1 to 4, fewer people were infected than when the count rose from 700 to 800.

None of the international platforms analyzing data on coronavirus infections exhibit a similar application of percentage growth rates. Through this confusing and misleading graph, the Armenian government is attempting to convince people that it is nearly “breaking the curve” of the pandemic. But it turns out that the formula published by the minister could work miracles on the statistics of countries where coronavirus is most widespread, such as Italy and Spain.

For example, if we were to hypothetically apply the methodology of Arsen Torosyan's published graph to Italy’s data, it would result in the conclusion that Italy has long ago “broken” the curve and recorded a decline trend.

The application of the percentage growth formula on Italy’s cases also clearly demonstrates that Torosyan’s formula cannot accurately represent any real positive dynamics.

Torosyan did not explain the purpose or methodology of the graph, but another Facebook user described it as evidence of how “Armenia is breaking the back of the infection.” By some coincidence, the unclear methodologies of users reporting this “breaking the back of the infection” narrative coincide with those used by Torosyan and the Armenian government. Incidentally, this is not the only instance of misrepresentation of declining trends regarding coronavirus cases prevalent in the Armenian information space recently.

As the number of infections increases and the geographical distribution of clusters expands, there are more attempts observed online to depict other factors in data analysis as a declining trend. However, the public needs information that reflects reality rather than generating positive spin.