Coming in April 2010

Isn’t it pretty? I’ll let you know when it’s available for pre-order.

Algorithm Fetishism or What Facebook Gets Wrong

My Facebook news feed is overflowing with people complaining about the changes to Facebook’s news feed/live feed. They boil down to two complaints (1) the news feed has too little and (2) the live feed has too much.

Facebook has also introduced a new form of suggestion — they are prodding us to “reconnect” with people, write on others’ walls, suggest profile pictures for those without them, and other forms of what can only be described as social meddling.This is causing much mockery amongst my peers as well as some horror — one friend reports that her friends have been urged to “reconnect” with a friend who passed away last summer.

At the core of the newsfeed and the suggestions problem is the same mistaken belief, one that I think all of these sites share (I’ve seen it so often on — the men who run these sites are so enamored of their algorithms that they trust them far more than they trust us to make decisions about what we want to see. People know what they want but Facebook is sure that it knows better. It’s one thing when Amazon or Netflix suggests a book or film I may not know about. It’s quite another when Facebook suggests a person with whom I already have a relationship.

If I were in charge of the site’s social engineering we would have checklists that allowed us to indicate whether or not we want to see who has friended whom, who has joined which group, applications, photos, notes, etc. We would have sliders to allow us to determine how much or which kinds of information we saw about which friends. The algorithms could learn from this, and adjust accordingly to make suggestions, but the end users would always have the power to override the algorithms’ suggestions.

Facebook’s new suggestions also make the mistake of thinking that the world of Facebook is the only world there is, so that if people haven’t communicated on site, they haven’t communicated at all. Thus I am being urged to reconnect with a colleague I see several times a week, a friend is being urged to reconnect with her boyfriend, and I am supposed to reconnect with someone with whom I am working on a grant proposal right now. My own research shows that on average friend pairs on use at least 2 other means of interacting. I would assume that number is higher for Facebook relationships.

I recognize (I’ve seen the data) that there is very little one-on-one communication that takes place through Facebook (or any other social network site), but there are good reasons for that, and it isn’t that we haven’t been reminded to do so. It’s that either we are communicating just fine off site, or we don’t want to communicate more than we already do. Thus the suggestions feel like either ignorant or invasive intrusions into our ability to manage our own social lives without their computer-generated wisdom guiding us.

I am just waiting for the sidebar suggestion: “Nancy, call your mother! She worries!” Because I’d never think of such a thing without them.

The 6 Types of Friends

I’ve been continuing to analyze the data I collected about friendships on Last week I presented a paper at Internet Research 10.0 in Milwaukee co-authored with Kiley Larson, Andrew Ledbetter, Michelle McCudden and Ryan Milner in which we combined quantitative analysis of motivations people had for friending with qualitative answers to questions about what they get out of friending. We then did a cluster analysis which led us to identify 6 types of friendships on the site. Axel Bruns did a wonderful job of live blogging the presentation and I hope he won’t mind my just quoting from his summary:

Nancy suggests that there are six types of friends: people who met on, divided into linkers, music explorers, and socialisers; people who met online, but not on (online socialisers); and local socialisers and local music socialisers.

Linkers have a static connection, very little communication, feel that it would have been rude not to friend, have the most recent friendships, and a low relational development; music explorers connect only because of the music, and have moderate and little off-site communication, they share musical tastes and histories, as well as other similarities, have the oldest friendship partners and low relational development; socialisers enjoy the site as a social space, do the most communication through it, met somewhere on the site, are interested but may not share one another’s musical taste, talk about music as well as other things, appreciate their differences, tend to be international and same-sex; online socialisers already knew one another from somewhere else online, and may also have met face-to-face, communicate a lot online but not through the site; local socialisers with high levels of face-to-face, phone, and online communication, but not through the site, observing one another’s listening and appreciate the sense of connection but don’t talk much about music, they have a moderately high relational development; and local music socialisers, who have the highest relational development, with high communication through all media,even moderately through the site, with music as a motivation for friending and an observation of each other’s listening patterns.

You can download my PowerPoints from the talk here.

I’ll also add that Axel blogged many other talks given at the conference, and point you to his complete event liveblog.