There were about seven birds of different colours that I used to see wobble around my neighbour’s urban chicken run. I frequently passed it on the sidewalk. Peering through the fence, I’d feel a bit of unease. I was a not-quite-vegetarian then: I’d given up some meat due to concerns about animal welfare, but still ate chicken. Yet here I was, charmed by these same feathered creatures.

What I felt was cognitive dissonance, a concept first described by the psychologist Leon Festinger in the 1950s. It’s the psychological discomfort someone experiences when what they know or believe is inconsistent with what they do. For instance, you might believe it’s wrong to lie but do it anyway, or know that a major corporation is harming people but still buy its stuff – and that might cause you some internal disharmony. Sometimes, people deal with that by rationalising their behaviour (eg, ‘everyone else does it’). Other times, they change.

A new paper by the researchers David Fechner and Sebastian Isbanner suggests that my increasing cognitive dissonance may have put me over the fence (so to speak) into vegetarianism. They compared several groups of people: those who ate meat and had no plans to change; those who were considering no longer eating meat; and those who had actually made the change. They found that cognitive dissonance (gauged by how uneasy, uncomfortable, etc one felt thinking about eating meat) was higher in the potential-vegetarian group than among the meat-eaters, and higher still among vegetarians.

Other factors differed too, such as how feasible a vegetarian diet seemed to them. But according to the researchers, the results suggest that cognitive dissonance helped to explain why some people who believed in vegetarianism’s benefits actually adopted it. When it comes to morally loaded behaviours, believing might not be enough. Our conflicts might have to stare us in the face.

My son was two and we had just been discharged from hospital following a nasty winter bug. His birthday was only a few days away with no time to plan for a big party, so I invited a close family friend and her kids to a teddy bear-making workshop followed by a pizza. She agreed, huge relief – the problem was solved. But then she cancelled last-minute because apparently her children had received a better offer from a popular classmate who hosted great parties! My friend realised quickly that her honesty had fallen flat with me and so she came over to apologise.

I was reminded of this event recently when I attended a lecture about the psychology of apologies by my colleague Shiri Lev-Ari. She described how research has shown that apologies are most convincing when they involve greater cost, such as in terms of money or time. My friend seemed to know this intuitively – she turned up on my doorstep (time cost) with a bottle of champagne (financial cost) at a time that would have likely inconvenienced her (effort cost).

Shiri wondered if this cost rule would extend to the words that we use when we apologise, and in her recent research that’s exactly what she found. People judged apologies involving longer words of explanation (I did not mean to respond in a confrontational manner) as more convincing than apologies involving shorter words (I did not mean to answer in a hostile way), presumably because they signal greater cognitive cost.

So, here’s my message to my friend: next time you need to apologise, do turn up with that that bottle of champagne but consider replacing your ‘real sorrow’ with ‘genuine remorse’.