home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!munnari.oz.au!uniwa!scott
- From: scott@psy.uwa.oz.au (Scott Fisher)
- Newsgroups: rec.audio
- Subject: Re: Very expensive cables
- Date: 15 Sep 1992 01:17:33 GMT
- Organization: The University of Westrn Australia
- Lines: 52
- Message-ID: <193djdINNqi6@uniwa.uwa.edu.au>
- References: <37994@sdcc12.ucsd.edu> <1992Sep14.171039.817@csc.canterbury.ac.nz> <19197kINNh6@uniwa.uwa.edu.au> <1992Sep14.153212.18854@scic.intel.com>
- NNTP-Posting-Host: wapsy.psy.uwa.oz.au
-
- sbradley@scic.intel.com (Seth Bradley) writes:
-
- >In article <19197kINNh6@uniwa.uwa.edu.au> scott@psy.uwa.oz.au (Scott Fisher) writes:
- >>Add to this a set of trials where you ask the subject to make
- >>subjective comments like "better, worse, same" to the manipulations
- >>and then we can not only start to see if there is any detectable
- >>difference but if the change is "better" and on what proportion of
- >>trials it was judged better.
-
- >Sorry, but you don't do subjective measurements in double blind tests,
- >it makes the methodology _a lot_ harder to get right, with much more
- >room for error. The way to do it is to randomly make A B assignments
- >in pairs (like AA, AB, BA, BB) with a single selection played through
- >each pairing. Then your panel simply needs to say "same or different".
- >Very simple, little margin for error. Better or worse comments can be
- >done informally (with the subjects still not knowing what the identification
- >is of A and B) later. This is how I've seen all audio double-blind tests
- >conducted, and that is how I'll be doing my double-blind interconnect
- >test.
-
- Notice I said, "add to this a set of trials", maybe I did not make myself
- clear. I was thinking about another experiment. The process I was thinking
- about is known as psychophysical scaling and in this case the equipment would
- be scaled (relative to each other) on a measure of "goodness" this should
- be done double blind. Some kind of paired comparison procedure sould suffice.
-
- To my knowledge I don't think any mag/rag or hi-fi manufacturer has tried
- a double blind, "goodness" scaling procedure.
-
- eg,
-
- Products A, B and C...
-
- present AB, AC and BC (randomizing the order within and between trials).
-
- Subjects hear two passages (same music, within trial, perhaps randomized
- between trials, perhaps not) simply respond (privately via button box)
- which sounded "best". After about 100 trials you should be able to
- rank A, B and C in order of "best-ness".
-
- Psychophysics is my living
-
- Regards Scott.
- _______________________________________________________________________________
- Scott Fisher [scott@psy.uwa.oz.au] PH: Aus [61] Perth (09) Local (380 3272).
- _--_|\ N
- Department of Psychology / \ W + E
- University of Western Australia. Perth --> *_.--._/ S
- Nedlands, 6009. PERTH, W.A. v
-
- *** ERROR 144 - REBOOT? is a registered trademark of ENSONIQ Corp ***
- -------------------------------------------------------------------------------
-