You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A very important ratio is missing: cases and deaths per million or 10,000 inhabitants
Here you see at first glance that countries with fewer inhabitants may be much more afflicted (deathsPer1mCapita) although the number of cases in other states (like US) is much higher as this result is ordered by that ratio.
It would be instructive to normalize this data with respect to a single country to get a feeling which country does better, e.g. with respect to de (having ratio 1.00)
You see here that Belgium has 9 times as many deaths per 1 million inhabitants than Germany. That number has more meaning than the ratio alone.
Next the ratio of cases and deaths is interesting. Well, we know that death numbers are not reliable (Corona may be interpreted as pneumonia) and case numbers are not reliable either (which people are tested, only those with severe symptoms or all infected?). But you get an idea nevertheless.
With the death-cases-ratio you see which country scores better (higher numbers are better here). At that time, Turkey did better than Germany and UK worse than Spain. Switzerland isn't much better than the US, and India follows directly behind.
Also, you see that for countries with a high population (especially India, China) the absolute numbers aren't helpful at all.
Next it would be interesting to see these ratios develop over time. This would give a clue whether the measures of different countries are working.
The text was updated successfully, but these errors were encountered:
Data is great, but it is hard to make sense of it. Some ideas to get a grip of what they mean...
Long lists are misleading and of little use. You want to compare only a small number of countries, for example
You might allow wildcards for countries you don't know the name of, e.g.
It is inconvenient and impractical to write country names. Use ISO codes instead or better alternatively
To give you an idea of how these numbers can be used to provide insight, here are some examples created with SQL. All results shown are based on https://data.europa.eu/euodp/de/data/dataset/covid-19-coronavirus-data from 2020-04-17.
A very important ratio is missing: cases and deaths per million or 10,000 inhabitants
It would be instructive to normalize this data with respect to a single country to get a feeling which country does better, e.g. with respect to de (having ratio 1.00)
Next the ratio of cases and deaths is interesting. Well, we know that death numbers are not reliable (Corona may be interpreted as pneumonia) and case numbers are not reliable either (which people are tested, only those with severe symptoms or all infected?). But you get an idea nevertheless.
Also, you see that for countries with a high population (especially India, China) the absolute numbers aren't helpful at all.
Next it would be interesting to see these ratios develop over time. This would give a clue whether the measures of different countries are working.
The text was updated successfully, but these errors were encountered: