The Personal Blog of Stephen Sekula

Sunday is under-count day

One of the things that often gets mentioned by coronavirus public information projects [1][2] is that certain days of the week are “under-reporting” days. Weekends tend to take the crown in this – public health reporting systems lag on weekends.

I can see this in the data from our county, Collin County in Texas. Plotting the average number of cases reported each day of the week (with the uncertainty provided by the error on the mean, or the standard deviation divided by the square-root of the number of data points), reveals the “weekend lag”.

The average number of cases reported per day of the week in Collin County. The uncertainty band (grey) is given by the error on the mean – dividing the standard deviation of cases by the square-root of the total number of data points in a bin.

The weekend is always a lull for us. The work week days are all generally compatible with each other within their uncertainties, but the weekends are indeed a suppressed outlier.

The lesson from the data: don’t take hope from the weekend.