COVID-19~ Should a weighted average be used to estimate r*?

We have used two ways to calculate r*, an estimate for r. We currently use the first (“Old”) method to estimate Ro for the early 2020 outbreak.

The only difference is on the second line where the average is used in the top calculation  a weighted average is used in the bottom calculation.

Note that n = 10 days and r[d] is a daily increase (growth factor) calculated over the next 5 days starting on day d for each country, and C[d] is the total number of Cases on Day d of the outbreak.

Originally and also most recently (including in the second half of last year 2023) we used an average (top calculation).

The two methodologies produced different values for r* and hence different rankings.

Note that we have adjusted the data for New Zealand since the images below were created.

Note: r# and r2 use the second weighted method; r1 uses the usual average.

r1, r2, and r# are all estimates for r*.

Old refers to the usual (also the currently used) average calculation.

New and Rank refer to a weighted average calculation.

Reminder: New Zealand is now between Denmark and Norway after data for New Zealand was updated late last year:

See:

COVID-19~ Better values for Ro

Shared Posts (Pingbacks)