[sspiff]

After reading the Nielsen' article [1] this page links to, I was wondering: this article keeps talking about "their user". Does that mean they tested this on one person, and then passed those findings as universal, or did they properly test this on dozens or hunderds of people?

I'm not contesting that carousels are bad (I'm not a web designer, but as a user I ignore most, except the one on the front page of Steam [2], but that may well be because I'm used to their biggest promotions and releases being there.) I'm just concerned about the quality of the test the Nielsen people did.

[1] http://www.nngroup.com/articles/auto-forwarding/

[2] http://steampowered.com

[dchest]

NN/g are well-known for developing methodologies for usability testing. They promote tests with only a few people, e.g. see http://www.nngroup.com/articles/why-you-only-need-to-test-wi...

[sspiff]

Thanks for the link! That explains their sample size. I don't necessarily agree with using a single tester though, the numbers in this post speak only of averages, which means you can still get a bad seed.