Enough here to keep me going for the weekend. Great post – thanks especially for pointing out the Scheiderman material.

So, basically, now people’s friends will be used as advertisements? Wow, we’re gonna become our own spammers….

Great post. The people we trust act as filters for all the noise that surrounds us. Friends and family know what matter to an individual. The good stuff tends to get through, the bad stuff stays blocked. If you didn’t have the different levels of trust, all the information could leave you overwhelmed and frustrated.

Spot on. Facebook is indeed doing this. I think it backfires.

Here’s my take on the circles that the perfect social network protects.

Huh, I’d have thought you’d find the Schneidermann diagram an oversimplification. If some peice of social software let you choose to mark other people on the network only as “Family or Friends”, “Colleagues & Neighbors”, or “Citizens & Markets”, I bet you’d criticize it as overdetermining relationships that are really much more fluid and complex than that.

The diagram doesn’t exactly say anything non-obvious, does it? And you could probably replace its assumed “Trust” with other social currencies and have it still make sense.

Does that mean that because I’m a friend of someone who loves PlasticFoodStorageContainers(tm) that I’ll now be subjected to adverts flavoured to attract them?

Doesn’t LinkedIn have a similar problem with the Friend of a Friend network growing…and no real relationship building other than the one to climb the career ladder?

Andrew…Schneiderman’s diagram is definitely an oversimplification. But I think that what is there isn’t wrong and can be helpful…in general we do ascribe more trust to friends and family and less to people we don’t know (citizens).

Where it gets interesting is where the model breaks down…for example we have authorities like Roger Ebert, for example, who we trust (well, lots of people do). And the question is, how do we model that in software?

In general, however, I agree with you. All diagrams of this sort suffer from oversimplification. The tyranny of each project rears its head…if each project didn’t have its own rats nests of issues then all of this would be easy.

Alex…thanks for the pointer to your diagram. I especially appreciate the “people like us” part, as I’ve recently been involved in a couple projects that leverage that. And, more importantly, I’ve heard several user interviews where that aspect has explicitly been called out.

It is over-simplication, and I have to respectfully disagree: yoummight trust your colleagues less, but I trust mine better for what they are expert in—e.g. I will never ask my mum about the latest band in town. More importantly, a colleague of yours has the same colleagues; it is hardly true half the time for cousins or friends — and tend to be awckward when talking about partners. Rather then measuring trust, you should realize you generally take what is available (scarcity of matching demands) and contextual framing: you don’t know what questions you should ask when buying a camera, but if your roommate keeps telling how great having several lenses is, you’ll probably go for a reflex.

Is this really new news though?? I thought this is old news, just rehashed for a new market – the Internet (blogs etc)

Bertil, you’re right. I was being general…there are definitely cases where you trust further sources more than closer ones.

In addition, the context of the situation matters greatly. For one thing, our decisions depend heavily on the physical spot we’re in. We might make completely different decisions if we at a mall vs. being at home with the Web.

So, let me qualify this diagram further with “in general, we trust those people we know more than those we don’t”.

Thanks for furthering this. I think as we find more and more ways to describe our decision making, we’ll build better software as a result.

For the record, it is “Ben Shneiderman” (no “c”). This is the 2nd most misspelled name in User experience, just after Jakob Nielsen.

Keith, thanks for pointing that out! I could have seen his name in a thousand places and never realized there wasn’t a “c” after the “s”. How embarrassing!