amsterdam_pete said:

It is worth asking what the claim would look like if it were false.

Worth saying that the confident version of this is more useful to the person writing it than to the person reading it.

The figures, for anyone assembling their own picture. If you are comparing against somebody else’s result, check that you are comparing the same measurement taken the same way. Most of the apparent contradictions in these threads dissolve at that step.

pete_manc_UK said:

Worth saying that the confident version of this is more useful to the person writing it than to the person reading it.

Adding the part of the answer the thread has not reached. There is a difference between no evidence and evidence of no effect, and this subject is one where the two get swapped freely in both directions.

Following on from KevinCompounds — and this may be the naive question:

How long did you give it before you decided it was working?

Reporting back.

Reporting back for the archive rather than for the thread: I did the narrow version of what was suggested, gave it the full interval, and it settled. Posting it because threads without a follow-up are half a thread.