I'm sure you've heard this before: "75% of knitters not only knit in the same yarn named in the pattern, they knit it in the SAME color as pictured in the book or magazine." I've heard it dozens of times. Naturally enough, with a varying, but invariably high, percentage named.
So how on earth do they collect this data? In theory, you could watch sales of a specific yarn after a pattern is published that uses that yarn AND track the cited yarn versus other colors. But for how long? For the two months until the next issue comes out? One year? Until the book goes out of print, during which time perhaps dozens more patterns using that yarn have come out?
I could also see gathering anecdotal evidence from yarn shop owners, "How often, would you say, do people knit things in the color shown in the pattern?" But I think this is inevitably skewed. Knitters who know what they want will just march in and get what they need. Knitters who need guidance will show the pattern to the yarn shop owner and get advice. The fact that they are less familiar with the yarns available and what they MIGHT use in a pattern, is, I would argue, clearly related to their likelihood to make a conservative choice.
Another by-the-wayside point: The color of yarn chosen by the designer had to do with trends and color availability in general. Maybe it's one of two "good" colors in that yarn. Perhaps robin's egg blue is what's hot this year. But many of the same things that made the designer choose that color will inform the knitters as well.
So, someone, please tell me where these statistics come from! Or, failing that, I propose an empirical test. We track a pattern, say published in Knitty, and the yarns used to knit it. I can't quite conceive of a way to make this a blind study, but when I look around at all the wonderfully various Clapotis (Clapoti? Clafoutis?) out there, I just think knitters are being sold short on their creativity and ability to make choices for themselves.
Either that or I can inflict my will on people more easily than I imagined :)