Axiom Home Page Forums General Discussion Stereo Sound & Vision mag reviews M60's

Double-blind is a catchy term but most people don't understand the strictness you have to adhere to properly conduct a true, controlled double-blind study. Controlled audio listening test are much more difficult and costly to setup and have many other variables that can introduce bias from both the experimenter and listening panel that almost make it a impossible to adhere to the double-blind standard (compared to the food example).

A true, double-blind listening test IMO you would need all these things.

Expertise: 3rd party (multiple) expert(s) in the field choosing the individual speakers and setting up the listening test and another 3rd party that is an expert in statistics and basic computer statistical analysis like linear regression/ANOVA that can take the data and perform and interpret the significance of the data.

Speaker Shuffler and acoustically transparent curtain: Adhering to the double-blind standard would be that you control any bias big or small and that would mean having all speakers being played in the exact same position in the room through a speaker shuffler and having the speakers not visible to the listeners.

Large random sample (100+) of people that we nothing about except having their hearing checked to get on the panel. They use their own source material, one at a time in the listening room, taking all the time they want.

The process would be expensive, time consuming and STILL be subject to plenty of error. That said, a controlled single blind test using random listening panels is more than appropriate for testing audio equipment. We are not talking about saving lives here.