Expectation bias generally relates to experimental design flaws, not (in this example) to subjects’ perceptions in a study. If expectation bias were involved in that study, it would likely be a fundamental issue rendering the investigators’ discussion misleading, as they’d be premised on spurious results.
This is an example of why it can be hard to discuss such topics in audiophilia - there’s a “minimum height requirement” in background knowledge for stuff like this to make sense among back-n-forth discussion (i.e. > 1 person). 😉 I know that looks reeediculously pretentious in print (Eew!) - I’m sorry, and I mean no offense. It just is what it is.