Was the D-Premier upgrade a waste of money?

[Jwg1749]

I think expectation bias plays a huge part, both ways. Without a doubt, it does for me. As is the case, I suspect, with just about all listeners and with just about all quality electronic audio equipment.

And, in my opinion, it goes hand in hand with why reviewers never submit themselves to robust blind tests (recognising as I do the imperfections of that process).

These things gladly keep much of the industry in place, and therefore lots of choice of big boys toys for us all.

I find the corollary of what you are saying rather depressing. It suggests we can't trust our own senses let alone the judgement of others. 

Guillaume

Well, you really can't trust your own senses. Your eyes have blindspots straight ahead. You just don't notice this because your brain fills in the missing parts. Your ears are easily deceived. Think about that strange noise you probably heard as a kid, that you thought was someone talking which turned out the wind through a branch or something like that. If you have a head cold you can't taste properly. Etc.

On the contrary, in normal operation our senses are pretty trustworthy. It's our minds that aren't. 

I found that, having accepted that expectation bias can play a very significant role in our judgements, I experienced a kind of liberation. I mean a liberation from the view that one has continuously to invest effort and money in improving one's system. I don't want to be stuck on that treadmill.

So my presumption is that tweaking electronics generally makes no difference. I'm happy to try tweaks out, if they can be explained rationally. But in general my knowledge of human psychology (I work on the margins of academic psychology) tells me that a null result is most likely to be the truth.

A good test case for expectation bias is the efficacy of psychopharmaceuticals. The market for anti-depressants is worth billions annually. And they evidently do make lots of people feel better. Yet the evidence from properly conducted trials is inconclusive.

Matt