On Twitter today I posted this question:
Why is Vermont’s death toll per confirmed case so much higher than in New Hampshire? Vt. is 13 out of 293 = 4.4%, NH is 3 out of 367 = 0.8%
Fortunately, I was smart enough to add this sentence: “However, the numbers are so small that it’s probably too early for any generalizations.”
It was quickly pointed by a Monitor colleague, Alyssa Dandrea, that one Vermont nursing home was hit very hard: Burlington Health and Rehab facility. Take away that one spot and the discrepancy disappears.
Small data sets can mislead us easily, especially if changes or relationships are expressed in percentages instead of absolutes.
nytimes.com, June 18. Deaths in Vermont: 55. New Hampshire: 330. Indeed, everything is relative and time changes everything.
In terms of the Spanish flue, almost a hundred years later it was claimed that the number of victims was exponentially higher than initially thought…