> A sample limited to accredited business and technology professionals who consider themselves early adopters and spend a lot on gadgets is, of course, not representative of iPhone purchasers generally.

1. Granurality: as OC mentioned, w/ an overall satisfaction "yes or no" method can reach there.

2. Bias: Pretty sure they did the research before releasing the product to get feedback from a loyal subset base/target customers.

Why would a study include non-Apple users if they have never used or refuse to use Apple products? Wouldn't this be biased if they are included to be "representative"?

I know people who don't use Apple products yet are vocal critics...humans can be highly ignorant creatures