1/ @justin_hart
BS FILTER: Facts vs Information

To filter out misinformation, one must be able to differentiate between the following:
A) on 6/1 there were 51 new deaths in OH
B) 2 people died on 6/1
C) 33 people died on 6/1

All of the above are true, but which are informative?
2/
Here is the daily new deaths chart for Ohio per The Covid Tracking Project. If you search daily deaths in Ohio on Google, this is the chart you will see:

It indicates that on 6/1, 51 new deaths were reported.
3/
The Ohio Dept of Health (ODH) provides an excel file that lists dates by date of death. If you download the file from 6/1, you will get a chart that looks like this:

It indicates that 2 people died on 6/1.
4/
If you download the ODH file from 7/14, you will get a chart that looks like this:

It indicates that 33 people died on 6/1.
5/
How can this be?

The answer is due to reporting lag. By the time an event occurs, it can take several weeks for it to be reported.

To compare the Daily Report with the actual without accounting for this lag is FACTUAL MISINFORMATION.

The "tail" effect can be seen here:
6/
Adding in the Daily Report (solid line) you cans see that the Daily Report significantly OVER and UNDER reports what actually happened on a given day.

The Daily Reports can be classified as FACTUAL MISINFORMATION.
7/
How to deal with this lag? Do we need to wait 2-4 weeks to get accurate information?

What is an ETHICAL way to report the facts?

The simplest is a 7-day moving average of the Daily Report, shown below.

While not perfect, it is a huge improvement over the Daily Report.
8/
An alternative is to look at weekly data.

Weekly data contains less "noise" due to daily fluctuations, allowing trends to be more clear.

The un-adjusted Weekly Report provides a fairly accurate representation of what actually occurred.
9/
In conclusion, just because something is a fact does not make it informative. Your BS Alarm should go off any time someone says something like, "I only care about the facts."

Good Problem Solvers clarify problems by verifying facts to produce information.
@threadreaderapp could you please unroll?
You can follow @dkwillabus.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: