You've distorted an ontological certainty into an epistemological compulsion. All effects have a cause. Not all causes are knowable. So it's one thing to say it happened, there must be an explanation (obviously true), and quite another thing to say it happened, let's assign an explanation (why?). To assign a cause without evidence is to sort among the sheep guts for a sign.
Let's not be silly. Katie changed her demeanor. It might because you yelled and it might be because she's preoccupied by holiday shopping. Since we have equal evidence for both theories (none.) neither can be asserted.
This is a long thread about nothing. Nobody has any clue why Katie's demeanor was different today.
Well, since we are striving for precision, I must point out you are wrong.
We do have evidence for scenario #1, since we observed her behavior and demeanor yesterday, and faxes, email, and telephone calls ensued (*evidence*) and today she appears with a toned down attitude.
Holiday shopping? You're right. No evidence she does her own.
Mind you, I do think there is something to the idea she was told to ratchet it down a bit thanks to the inundation of irate viewers' complaints, but this post is addressing your incorrect assertion that we have "no evidence" as to at least one reason she behaved differently today.
And I'll just add that we have presented other instances where the audience outcry has resulted in changes. I am fond of my Arnett example.