CIA now says COVID-19 'more likely' to have come from lab

[Igel]

18. Think of it as on an x-y plane.

One axis says how reliable your data are.

The other axis says what the data say.

Think of this as the spatial equivalent of "intersectional", with only two parameters.

You can have rock-solid data that really doesn't answer the question one way or the other. Or you can have weak data that points clearly in one direction. Sometimes weak data are all you have, and "weak" doesn't mean "wrong." Just you can't trust them to be right (or wrong).

The IPCC reports are the same way. If you don't report *both* bits of information I think it properly counts as misinformation, and very few news sources report both bits.

Always remembering that even the evaluation of the data as low or high confidence has confidence intervals associated with it.