Just to add Frantz, this is a good example of what has been described as bias within A-B comparison.
It is scientifically known that if a listener is unsure of differences then they will instinctively guess on a choice, and was identified in most cases majority will go with A.
This is partially resolved by creating force choice situation, where the listener chooses and also states the difference, or chooses and states unsure or even cannot tell a difference.
This is a very simplistic example to the AB selection bias effect, the key IMO is self perception and understanding the mechanism in action, which can help.
In this case if the listener cannot differentiate then they need to overcome the heuristic-cognitive decision process that makes them choose A or B, so the answer would be cannot tell a difference.
Other factors-biases include that when we think something has changed even if we think they may sound the same we start to think they will be different, due to our thinking that a change has an affect.
There are a few others that we touched upon in a different thread in ABX on this very subject, think it was discussing it with Ethan.
But interestingly there is an example of a cable listen test on another forum-review publication where one did overcome the false change (the other 3 failed to note there was no difference while the one who passed said he could not tell a difference and so thought he failed).
Not bothering to link as what the test proves has to be taken as limited, such as this example.
Cheers
Orb
