The dbt is very popular in other areas of research like long term medical experiments, to avoid that the researcher himself is influencing things. That is not bomb proof either and is just as important afte the test as the data can be consuisly or unconsiusly doctored after the fact in the statistical treatment.
So the verdict betwwen a and b is ideally decided without testpersons or researcher have any clue what is what. But even then it can be botched, the reasercher can filter out " anomalies" or other odd data that does not " fit " . This does not need to be deliberate , the researcher can feel strongly for one kind of outcomme and skew the dataset. That level of rigour may not be needed at home :) However what may be needed is enogh test, I forgot most of my maths sorry . Q what is statiscally signifant ? I get that if 3 of 5 people prefere a vs b it does not prove a thing as it could easilly be a coincidence , for so few tests would not 5 of 5 be the only reliable result ? So a larger dataset is needed. Likewise results like 52/48 in large dataset may be unreliable to. Is it any accepted easy formulla that is " good enough" for our purpose . For example if you do 10 test or use 10 test subjects how good must the verdict a vs b be to be trusted ? 6/4 or 7/3 . But if you are a bunch of people it would be doable to have everyone do the test 10 times or so. If you do it all by yourself you migth be very bored trying to take 100 test :-/ -- Mnyb -------------------------------------------------------------------- Main hifi: Touch + CIA PS +MeridianG68J MeridianHD621 MeridianG98DH 2 x MeridianDSP5200 MeridianDSP5200HC 2 xMeridianDSP3100 +Rel Stadium 3 sub. Bedroom/Office: Boom Kitchen: Touch + powered Fostex PM0.4 Misc use: Radio (with battery) iPad 64gB wifi +3g with iPengHD & SqueezePad (in storage SB3, reciever ,controller ) ------------------------------------------------------------------------ Mnyb's Profile: http://forums.slimdevices.com/member.php?userid=4143 View this thread: http://forums.slimdevices.com/showthread.php?t=91322 _______________________________________________ audiophiles mailing list [email protected] http://lists.slimdevices.com/mailman/listinfo/audiophiles
