I wrote to them recently concerning their testing procedure for audio
electronics. I don't subscribe, although I occasionally leaf through
library copies when I'm supposed to be working :-)

In answer to the Q: do you use blind testing and matched levels, she
said 'yes'. However, she went on to say that it's not always needed
because, as they're completely independent, there is nothing to gain
from recommending any particular model. She also went on to say that the
tests are thorough to replicate several years' use.

I thnk this is generally fine, and nice of them to respond. It's also
rather vague - I think I'll reply asking them to publish the detail of
their tests (level matching, respondent selection, room, other kit,
music and medium used), and suggest that they think about:

Double blind testing, or at least blind testing - they may be
independent, but people aren't, and some may think a Sony is bound to
sound better than an Alba. Sort of reassuringly, their tests rarely
mirror the specialist press (at least on the occasions I've been able to
compare the 2), but I think blind testing might add some rigour.

I think that would do me - I don't really want to know about cables, and
while some technical data (power output maybe) might be useful, I don't
think they have the expertise to measure/interpret the information. Just
knowing that people can reliably and qualitatively differentiate is a
pretty good start IMO.

Just FYI, but if anyone has any comments I'd be interested.

Rob