[scottcowley]

Related to this, there is probably a psychological effect at play where people want to be consistent with what they have said in the past (see Cialdini and consistency theories), so just the act of asking the research participants to articulate their feelings about the person dug themselves into a hole. Talk less and you don't have to entrench.

[kelukelugames]

I apologized in a meeting yesterday when I realized I said something factually incorrect. I actually fact checked myself to discover the error. I wonder if that was a mistake. Given how our leaders act, I think we are more wired to trust people who never acknowledge their mistakes.