Hacker News new | past | comments | ask | show | jobs | submit login
34.5% of US Internet Population Not using Facebook/Twitter (jamiegrove.com)
106 points by hownottowrite on Sept 9, 2012 | hide | past | favorite | 42 comments

"There are three kinds of lies: lies, damned lies and statistics."

And then the OP runs a questionable survey that doesn't warrant the conclusion (which by the way, is not earth-shattering).

- You could not be on Facebook / Twitter and still pick one of the other answers.

- You could be one of the services but not the other, and pick the first answer. What does Facebook "/" Twitter mean? AND or OR? Completely unclear.

- If the order of the options wasn't randomized, you would likely get a bias towards the first option. Don't know if this was the case.

The article does say the order was randomized.

But there is a bigger problem: one of the options is "I’m scared of scams." Regardless of reading order, if you read all the options before answering - and presumably most people do - then this causes what psychologists call a "priming" effect. You start to think about scams, and you can then behave very differently than if that option were not in the survey.

I agree. There's a great "Yes Prime Minister" video about this. I just wrote a blog post about how it relates to this poll.


Perhaps I missed it, but that was missing the answer I would have given, which is "No - Because I don't trust any company to give them that much access to me, even if I do like their service". The closest thing was "No - I'm afraid of scams"; but I'm not really afraid of scams as much as lost/misplaced data, company policies changing with respect to my privacy, etc.

So the scam you'd be afraid of is Facebook or Twitter.

Agreed; but I'd put that (missing) option as: No - due to privacy concerns

This. It's not (necessarily) a matter of scam, I just want to keep my activity in different parts of the internet segregated.

Yeah, that's a lot more concise.

There's not an answer on there I could honestly choose if I'd taken that survey. I would have voted No. But I am on both services, I do understand how it works, and I am not scared of scams.

I just don't see why I should - creating an account using an email address is also easy and by this point don't most people understand how it works? If they managed to sign up for facebook or twitter, they clearly do!

creating an account using an email address is also easy

How about companies like Pinterest and Spotify that only let you sign up with a Facebook or Twitter account?

One of the most interesting part of this article is that it made me discover this google survey tool.

For the rest, I'm quite sure there's a skew here, because people without internet access or not using it for browsing (mostly older people) should be under-represented. I don't want to believe that 65% of everybody in the US is using FB.

It's somewhat pointless to ask people without internet access if they're interested in logging into websites using another website account.

So yes, that's the web using subset of total population, totally fine for this question.

That aspect is more of an issue if you want to use Google Surveys to determine whether to start a new supermarket on Main Street. I hope the Google Survey people propose some other service for such inquiries.

It said pretty clearly in a couple of places that it was a survey of internet users.

I wouldn't sign up for a site using FB/Twitter unless there was a solid technical reason to (i.e. Buffer and Twitter, because it's a Twitter-posting app).

And that's because I see no reason to associate my Facebook (family/friend-facing, lots of personal stuff on there) with any company. I like to keep a strict firewall in there.

>Google has a whitepaper explaining how this tool works and why it’s as good if not better than doing telephone surveys. I’m going to trust that Google’s stats wizards know their stuff

But... they have a clear bias in favor of telling you that it's better than telephone surveys. Clearly you should not trust a claim like that without looking at an independent analysis.

Missing option: no, because I hate it.

That's essentially my answer. I have accounts with both Facebook and Twitter. I understand how the authentication works, and I'm not afraid such requests are scams (note: that's mostly a subset of not understanding how it works). I don't want my online identity tied to a third party account whose operator has few obligations to me.

Yes, this survey design is questionable by most standards.

But, it's not a survey design disaster and is likely a "directionally correct" reflection of internet users' actual behavior.

Are exactly 34.5% of the US Internet Population not using fb/twitter? Of course not, but I suspect that number is pretty close.

If you're building a social networking app, this is an opportunity.

The result looks reassuring, that there are still people who care about their privacy. (I'm not using Facebook and Twitter both).

Requiring to sign in with FB or Twitter is disrespectful. If some site offers external authentication, let it use OpenID or BrowserID already.

Wouldn't a better title be "65.5% of US Internet Population using Facebook/Twitter"?

Interesting results, but I'm skeptical as to whether the sample results can be extrapolated to the U.S. internet population as a whole.

That also means 65.5% of US Internet population is using Facebook or Twitter, a mind boggling statistic

65.5% of the web browsing, any-other-demographic-and-psychographic-modifying-aspect-of-Google-Survey population. That makes it harder to say how boggling of a statistic it is. Far greater than 65.5% of my household accesses Facebook and/or Twitter.

"Far greater than 65.5% of my household accesses Facebook and/or Twitter."

In my social circles, Facebook/Twitter usage clocks in at about 15%, so the numbers of the survey were at least minimally surprising ...

I would wager that a certain percentage of people in the US have no Internet service at all. Your guess on that percentage is as good as mine. I can think of several older people I know that do not have Internet service or cell phones. Nor do they care to.

I wrote "65.5% of US Internet population", not "65.5% of US Residents".

There's hope for humanity.

1,817 random samples is not a good way to represent a population of 300,000,000+. What kind of audience is represented by someone who would take a random poll on the internet? Probably someone who is not that tech savvy, skewing the already small sample size.

1,817 truly randomly sampled people would be pretty great and plenty for good data.

The problem here is that the sample is not random.

Sometimes I let sites sign me in with FB/Twitter, because the alternative is giving them my email address, which seems nearly just as bad. And that way, I don't have to type anything.

There's easily an infinite amount of email addresses (eg. spamgourmet). A browser extensions to automate that would be nice, though.

FB and twitter both spread enough tracking beacons (those like/share buttons) across the web that any attempt to use >1 accounts to distract FB or twitter are probably futile.

I always use a unique, disposable email address for every account I create, and use KeePass for automated login. This method is just as easy, and allows me to retain full control over my online accounts.

It's also a fun way to see who sells your email address to third parties.

Interesting that the percentage of a group that claims not to be on FB/TW goes from

* a high of 41.7% for the 65+ group.

and moves steadily down through the groups to

* 27.4% for the 25-34 group.

but then jumps back up to

* 34.7% for 18-24 year olds.

My theory: Twitter with a pay-as-you-text phone plan is not wise (if you like to get tweets as texts). Hence the kids can't afford it.

Do most kids and people in general have all you can eat text or pay as you text? I've been pay as you text, but now that my teenage son has his own phone (and all you can eat text, because he's gonna text ...) I will have to move my plan to all you can eat just to survive his comms.

Interesting survey but I think an A/B test is really what we need here. A = traditional authentication / B = FB, Twitter, Google OAuth.

The people without FB and Twitter are late adopters and won't try your service no matter what kind of auth you use. For the people that are concerned about privacy, I'm not sure how big their impact will be.

I am not a late adopter. I do have accounts on Facebook and Twitter. I will not sign up for your service if I have to use one of those accounts to do it because I don't want either one of them becoming my identity online.

This should read like all other surveys:

34.5% of the people who took this survey claim they are not using Facebook/Twitter.

I agree. The amount of response bias in this survey will tend to be absurd. I suspect most people who use facebook and/or twitter won't bother with answering this. There is no good reason to. I also suspect that people who don't use facebook or twitter will also be indignant about the fact that they don't use it and will make a point to vote against them. Heck even I have done similar stuff before. From the outset this survey isn't going to work if the methodology is not correct

Average IQ has gone 34.5% up too. What a coincidence.

The average IQ is, and always will be, 100. By it's very definition.

Thank God.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact