The three letters DBT are the refuge for those who were never able to fully describe valid blind tests and want to write short posts.
IMHO, the main question is not the method but its implementation and aim. Any correct method to avoid external bias is welcome, if it does not affect the test it self. Except for the tests carried by the Harman group or papers from research groups studying psychoacoustics and/or compression algorithms, I have never read about a proper documented audio blind test and its analysis.
Can any WBF member describe with detail a properly carried blind test in which
he has participated, its statistical analysis and conclusions?