The fundamental attribution error
Researchers have tested this many times, in different places, over many years, and they keep finding the same result. This one is safe to trust.
Two psychologists, Edward Jones and Richard Nisbett, laid out a basic asymmetry in how we explain behaviour. When someone else does something, we tend to explain it by their character: he was late because he is careless, she snapped because she is difficult. When we do the exact same thing, we explain it by the situation: I was late because traffic was terrible, I snapped because it had been a brutal week. We judge others by their actions and ourselves by our circumstances. Few patterns in social psychology have been studied more, and the later checking found the split strongest exactly where it does damage, in explaining behaviour that lands badly, which gives it a sharp edge in any setting where we judge a person from a thin slice of behaviour. An interview is exactly that: the single most unusual, decontextualized hour of a person's life, read entirely as evidence of who they are.
You explain other people's behaviour by their character and your own by your situation, and the gap distorts every judgment you make of a person you do not know well. The candidate who stumbled is a poor communicator, but when you stumbled yesterday it was a hard day. In an interview, a person is nervous, out of context, having their worst or strangest hour read as their permanent self. Before you code someone by a single moment, ask what situation might produce that moment in a capable person, because for yourself you would grant that situation instantly.
Solomon's paradox says you reason better about a friend's problem than your own, because distance strips away the fear and ego that cloud your own decisions. The fundamental attribution error says you explain a stranger's stumble by their character and your own by your circumstances. Same distance, opposite effect. Distance improves your thinking when you are solving and degrades it when you are judging. So the advice you give a friend is wiser than what you do yourself, and the verdict you pass on a candidate's worst hour is harsher than the one you pass on your own. The useful move runs in both directions: borrow the distance when it is your decision, and lend the context when it is someone else's.
Source: The self and other split: Jones and Nisbett, 1971, The Actor and the Observer. The naming of the wider error: Ross, 1977, Advances in Experimental Social Psychology. The modern check: Malle, 2006, Psychological Bulletin, a meta-analysis of 173 studies finding the split holds mainly for behaviour that lands badly.
Ross and Nisbett, 1991
On how much of what we call character is actually the circumstance a person is standing in.
Draw your own card. It does not take long, and it rewards taking your time.