In a recent study of arthritis, researchers tried but failed to find any correlation between pain intensity and any of those features of the weather—humidity, temperature swings, barometric pressure—usually cited by arthritis sufferers as the cause of their increased pain.
Facts
A study couldn't find any correlation between weather events and arthritic pain intensity.
The people who were convinced there is a correlation gave very different stories about how it works (does the weather event immediately cause pain? does it cause it hours later? days later?)
Evaluate
If we're trying to pull these strands together, there's a tension between people claiming there is a correlation and scientists being unable to find any correlation.
Do we know which way to lean?
If there really is no correlation, then why are people claiming there is? Maybe it's selective memory and once they started thinking there was a connection they had confirmation bias (remembering strongly the times when the connection held and forgetting the times it didn't).
If the people are correct that there is a connection, why couldn't scientists find any correlation? This one is harder to explain. We'd have to say, . That's already a little corny. But we'd also still probably find some correlation between rain and joint pain if that were true. It might be a weak correlation because of the varying onset of symptoms, but it wouldn't be NO correlation at all.
Goal
It seems like we should trust the data over the individual people's testimonials. If we can't find any correlation between X and Y, then X probably doesn't cause Y. That would mean that these people's beliefs are wrong. Let's look for a safely worded version of that.