Yesterday, in our post on perfect pitch (usually called absolute pitch in research reports), we offered a quick test to see if we could identify the portion of our readers with absolute pitch. At first, things were looking good for the absolute pitch crowd. Readers listened to this note:
A whopping 18.8 percent of the 165 respondents identified it correctly as E. Since random chance would predict that just 8.33 percent of responses would guess this note, it would appear that over 10 percent of our readers have absolute pitch.
But some readers pointed out that many string instruments have an E string; this is a particularly easy note to guess — that’s a good point. So we changed the note and conducted a new poll. This time readers heard this:
Only 7.2 percent of 195 respondents correctly identified the note as G-sharp — a rate lower than the 8.33 percent random chance level. What’s up with that? Are there *no* readers with absolute pitch?
There is a potential explanation. Athos et al.’s study noted that even people with absolute pitch tend to make a disproportionate number of errors on “black keys” — sharps and flats. So I computed the average number of responses on white keys and black keys. On average, people guessed a given white key 10 percent of the time, while selecting a given black key just 4.5 percent of the time. Now we can compare our actual results to these baseline figures:
So in Poll 1, people correctly identified E 8.8 percent more often than they selected other white keys, and in Poll 2, people correctly identified G sharp 2.7 percent more often than they selected other black keys.
Now, there are some other problems with our data. One commenter who claimed to have perfect pitch identified the G sharp as an A. This is consistent with Athos et al.’s finding that people with perfect pitch often misidentified that note, possibly since orchestras tune to a wide range of As. Indeed, in Poll 2, 12.3 percent of respondents said the note was an A, 2.3 percent more than chance, and 2.6 percent more than in Poll 1. It’s possible that that extra 2.3 percent includes many people with absolute pitch
So can we say that at least 2.7 percent, and up to 5 percent (2.7 + 2.3) of our readers have absolute pitch? Actually, that figure only represents about 10 individuals, so we certainly can’t say it with precision. I should also point out that during the time the poll was up, around 2,000 people visited the site, but only 360 responded. Maybe they didn’t respond because they had absolutely no clue what the answer was. So it’s possible that the actual percentage is substantially lower. Statistical error rates mean the percentage could be much higher as well.