Flutterby™! : Double-Blind Strad

Next unread comment / Catchup all unread comments User Account Info | Logout | XML/Pilot/etc versions | Long version (with comments) | Weblog archives | Site Map | | Browse Topics

Double-Blind Strad

2012-01-09 17:59:16.35984+00 by Dan Lyke 1 comments

Remember that study which supposedly claimed that professional violinists couldn't tell a Stradivarius from a modern violin? Not so much: They were asked which one they preferred, and were never asked to identify which one was the Strad. One of the participants goes into detail:

I was not asked to identify specifically which was the modern violin and which was the old violin; only which I preferred. If people are concluding from this study that "professional violinists can't tell the difference between modern violinist and old Italians," then I think we need a different study in which violinists are actually asked to identify that.

[ related topics: Interactive Drama Music Weblogs Handicaps & Disabilities ]

comments in ascending chronological order (reverse):

#Comment Re: made: 2012-01-09 21:04:52.413411+00 by: ebradway

Pretty typical media drawing illogical conclusions from an experiment. It's easy to forget that experimental design typically disproves a null hypothesis. That is, it only shows that one possible outcome is not likely. This particular study showed clearly that the professional violinists did not prefer one particular Stradivarius over the other selections, which included some modern violins. Without looking a the publication, I suspect the study's results conclusively showed very little else.

Of course, the experiment had a design flaw in that the Strads were tuned like the other instruments. But I suspect that all of the violinists were instantly aware of not only the fact that those instruments weren't tuned quite right, but also that they were probably Strads.

The media response reminds me of an article I recently read where the author said something along the lines of "the results were inconclusive, showing X was barely greater than Y but only within the margin of error" but later stated "X and Y were different with X being greater." The margin of error means that if the experiment were run again, it is very likely that Y would be greater than X by a similar amount. Can't do that, guys!