An analysis of my favourite performances against the piano selection (WARNING: Geeky stuff)
It occurred to me that knowing the choice of the piano may have influenced my view of the performances. So, I have done a little analysis of my evaluation of the competitors (0 - 3) against the pianos (1, 2, 3). I used the following scale for the competitors:
- 0 = I don’t want to hear this person again.
- 1 = This person could advance
- 2 = This person should advance
- 3 = This person will advance
And the pianos are:
- 1 = New York (my least favourite)
- 2 = Van Cliburn Foundation New York
- 3 = Van Cliburn Foundation Hamburg (my most favourite)
My scoring of the competitors had a mean of 1.7 with a standard deviation of 0.8.
The correlation between the choice of the piano and my evaluation of the performance is -0.04, which means that the choice of the piano has no relationship at all to how I perceived the performance.
So, I was relieved to hear that I was probably hearing the music, not the piano.
Ken Iisaka
May 26th, 2009 at 6:47 pm
It’s difficult for me to put “numbers” to the performers since they played different pieces of music and it’s like comparing apples and oranges.
However, there are a few performers, regardless of the repertoire they chose, that really shine.
Spencer Myer, Eduard Kunz, Lukas Vondracek, Evgeni Bozhanov, Stephen Beus, Ran Dank and Mariangela Vactello.
I really want to hear them play more.
May 26th, 2009 at 6:52 pm
OK. Now please correlate with your opinions of wardrobe and facial expressions.
(Ken, I think your comments about the pianos have been interesting.)
May 26th, 2009 at 6:59 pm
Unfortunately, I think that some notes were out of tune on the upper register during several performances. It reminded me of Joyce Yang’s performance during the final in the last competition when it sounded as if a string (it might have been F) got knocked out of tune. I remember hearing it during her Schubert-Liszt and Brahms. This time, during Vondracek’s Chopin Nocturnes I heard several notes (that may have included the Eb) sound amiss. Also, during a lovely performance of the Mozart Piano Sonata by Lifits (or possibly Bozhanov) I could hear the D and several other notes were out. (Please feel free to specify the exact keys.)
May 26th, 2009 at 7:18 pm
OK. Now please correlate for your opinions of wardrobe choices and facial expressions.
(Ken, I’ve enjoyed your remarks about the pianos.)
May 26th, 2009 at 7:31 pm
her playing is outstanding,her music has speak and touch to me
May 26th, 2009 at 7:48 pm
I’m not sure you can use a correlation coefficient with categorical vs. ordinal. I believe you would want to perform an ANOVA or, even better, the small sample alternative.
May 26th, 2009 at 8:45 pm
JBB is right: correlations are only for interval-scaled data. For ordinal data, one needs a more complex measure, like Tschuprow’s. But now it’s getting REALLY nerdy.
May 26th, 2009 at 9:42 pm
Ken’s “geeky stuff” had me thinking of the geeky new forumula’s the ISU came up with for Olympic Figure Skating. :- ) Concerning who should move on- There were simply too many wonderful piansts that we have heard over the past 5 days for only 12 to advance. I VERY much want to hear at least 18-20 of them again and know I am going to be extremely disappointed in a short while. Thank goodness for the Preliminary Archives. Kudos to all the staff and folks who made all this possible online.
May 26th, 2009 at 9:42 pm
StatsProf, et al,
You are absolutely right. I was thinking of Spearman’s…
May 26th, 2009 at 11:07 pm
I’d be interested in looking at the piano data, at least, as I suspect there will be a relationship. The Hamburg sounded so much better than the other two.
May 27th, 2009 at 12:14 pm
I took a few minutes and compiled the data myself. Here is the distribution of semi-finalists versus non-advancing versus piano:
Not Advanced Advanced
Hamburg 6 7
NY Gloss 6 3
NY Satin 5 2
Those advancing overwhelmingly chose the Hamburg, whereas the non-advancing pianists more or less chose amongst the other pianos randomly. Amongst those who advanced, chosing the Hamburg was statistically significant. However, there is no statistical significance between advancing and piano choice (hope this makes sense, read the two statements carefully) as many who chose the Hamburg didn’t advance.