Confirmation bias works both ways. You do something that you don't think makes a difference, and so you don't taste a difference.
That is the problem with confirmation bias. One of the reasons why I posted my "Shaken, not Stirred" method was to see how well it stood the test of time with other brewers. I knew that it worked for me, but was it conformation bias, or did the method provide an easier, effective way to make a starter? At this point, I am fairly certain that what I thought may have been confirmation bias is actually real. I think that as a whole, we need to be open to testing what people put forth. That is how science works; namely, does the experiment hold when conducted by other people? I am a firm believer that everything matters with delicately-flavored beers.