Hacker News Reader

Comment by mvanveen | Hacker News Reader

Given the overwhelming sample size of 37, I think we can all agree that this is statistically significant information which one should base big platform decisions on.

View on HN

You might have missed the second half of the post:

> This is a better test, because the signup screen offers both options with equal weight and we have over 70,000 organic signups.

Still wasted the first half of the article talking about the results of a 37 person survey. 37 people with no demographics info at all is so anecdotal that it's effectively worthless.

As far as the second test goes, of course the general trend will be younger people who actually use Facebook will not mind... using Facebook.

I don't think that's true at all. Useability guru and generally big-headed man Jakob Neilson says you can get the majority of content from user testing as few as 5 people. Granted this isn't user testing, but small sample sizes are not pointless, they're directional.

If people were indifferent (50% chance of wanting facebook) then the likelihood of at least 27 out of 37 randomly asked people saying "no facebook" is about 1 in 300. And that's counting the "Yes but I hate it"s as yesses. Count those as "no" and it's 1 in 3000. Highly significant.

This assumes the "random tourists in Pike's Place market" are an effectively random sample. It's possible that facebook fans are less likely to take vacations in Seattle, but Occam's razor is against it.

Reply on news.ycombinator.com