Comments on: Most fans paid $0 for Radiohead album – Yahoo! News

By: Nicole

Nicole — Sun, 11 Nov 2007 23:42:42 +0000

@Stuart: Thanks, but I understand statistics. However, I have a problem with the way this information was reported — no mention of who ComScore is, how they compile their database (which could already be biased data), or why their information should be trusted, in addition to a lack of other information like margins of error. When you’re talking about a sample size of “a few hundred” margins of error can be huge. And about that sample, “a few hundred” doesn’t seem like a sample size that would result in a proper representation of the entire world population.

By: Stuart Miller

Stuart Miller — Sat, 10 Nov 2007 12:40:57 +0000

I don’t know what the fuss is all about. It’s called sampling theory and it’s been around for years — and it works. If you don’t believe in the validity of sampling, next time you go to the doctor and he wants to “take a little blood”, you should tell him to “take it all”!

By: Jason Sares

Jason Sares — Wed, 07 Nov 2007 17:15:46 +0000

The problem is journalists come up with an angle and then look for sources that support their idea.

By: Nicole

Nicole — Wed, 07 Nov 2007 15:34:29 +0000

Wow. That’s horrible.

Most #1 releases in CD format in the US sell 100k – 200k in the first week. Let’s say conservativley that 100,000 people downloaded the Radiohead album in an entire *month*. So they surveyed “a few hundred” — we’ll say that’s 400 — leaving us a sample size of .4%. Perhaps not horrible by itself, but ComScore has a database of only 2 million users *worldwide*. Thus, we’re working with a sample of a sample of a population. There’s also no mention of margins of error.

Veiga apparently has no taste for statistical follow through.