3 Mar 2010

Facebook protest groups - from lazy statistics to lazy journalists

We've all seen them. If you have a few hundred friends on facebook then the chances are that every time you log in, you will discover that one of your friends has joined a group that is protesting for or against something. Many of these groups are named after a specific intention to get X members to join. E.g.

'I bet that we can get 1 million facebook members to put an end to domestic refrigerator abuse'

I will say from the outset that I don't have any particular problem with the concept of having these groups on facebook. There are many just causes in the world, and facebook is an easy way of spreading your message about these causes. Not only is it easy to do, you also have the possibility of attracting many, many people to your cause. Instead, my issue with these groups is two-fold:

1) Many of the groups that attempt to get a target number of people to join their cause never consider whether that is an appropriate target or not

2) Journalists seem to love including a piece at the end of articles saying "There is now a facebook group in support of this protest which has attracted X many members".

Let's explore each of these issues in turn through use of some specific examples.

Same sex marriage

Here is a typical facebook group to illustrate my point (I include the current membership of the group in parentheses):

If you are going to pick a target number for your facebook group then '1 million' seems to the most popular choice. It may be a convenient 'very large number' to choose but it becomes a little bit pointless if your target audience already exceeds that figure. I.e. if I start a group called 'Lets get 1 person in my household to not burn the house down', it's not very impressive if I find one person to join.

The potential membership of any facebook group is theoretically limited to however many people are on facebook.  There can't be that many people using facebook, can there? Umm, yes there can. Facebook's own statistics puts the number of active users at 400 million. It doesn't break down these statistics by country but it does say that 70% of its users are from outside the United States. This implies that there are ~120 million facebook users in the United States, which is about 40% of the entire American population.

So what does all this mean? Well, for the group mentioned above it means that the group owner is trying to get 1 million users out of a potential USA facebook population of 120 million. If we ignore any interest in this group from people outside of America then it is essentially asking: 'Can we get 0.8% of Americans to join this group'. Whether this is an impressive feat or not partly depends on how many gay, lesbian, and bisexual people there are in America. It seems safe to assume that most, if not all, of these groups should be in favor of same-sex marriage. According to Wikipedia, a 2008 CNN exit poll suggested that self-identified gay, lesbian, and bisexual voters made up 4% of the voting population. Other surveys have put this at higher proportion, but let's go with 4%. By proxy, we'll assume that 4% of American facebook users are gay, lesbian, or bisexual. This suggests that there are 4.8 million gay, lesbian, and bisexual Americans who use facebook. And yet, the group that is campaigning for same-sex marriage has attracted less than a third of this number. Suddenly the 1.6 million membership of this group, which surely includes many heterosexual facebook users, looks less impressive.

State vs state

This is a little more interesting. I assume that there are similar groups for all states in America. You hopefully are already aware that California and Texas are the largest and second largest states by population. You may not know that Vermont is the least populous (I'm excluding Washington DC). We might assume that there are no huge state-wide differences in the use of facebook, but there are some figures that dispute this. California is 1.5 times more populous than Texas, and yet it has nearly twice the number of facebook users (the excess Californian facebook users are no doubt concentrated in the Bay Area). However, if you calculate each of these groups' memberships as a proportion of all the facebook users in that state, then we see a different picture:

Vermont - 24%
Texas - 19%
California - 2%

Well done Vermont, you're punching above your weight!

Feel the love

This example really doesn't need much discussion. For one thing, do you think that many people hate planet earth? Assuming that the answer to this question is 'no', then we can see that this group is targeting 1 million people which is a quarter of one percent of all the world's facebook users. That's aiming pretty low, but as the group membership shows, it's not quite low enough.

Aiming high

Hooray, at last a group with some degree of ambition. Even better than just targeting a much larger number of people, the owner of this group deliberately targeted a proportion of all facebook users worldwide. The problem with this approach of course, is that things change quickly. At the time of creation, the group was aiming for 55 million people which represented over 90% of all facebook users. It now represents about 14% and this proportion will only continue to fall. But at least it is along the right lines.

Facebook you can stop this madness

Ideally I'd love it if facebook itself provided a way of creating group names where you can specify a target percentage and let facebook update the target figure based on its changing membership. In my fantasy world, facebook would even give options where you could specify a percentage of the world's facebook user base, or just a specific country. Of course, it's unlikely to happen. But until that day comes, I live in hope that more people start realizing that targeting a million people may not be that ambitious at all. The degree of ambition does also depend on the cause however. E.g. which of these three (fictional) causes are more deserving?

1 million people to abolish fox hunting
1 million people to abolish Fox News
1 million people to abolish Megan Fox

Extra, extra, read all about it

Speaking of news leads me to the second part of my argument against facebook groups. Most of these groups are not newsworthy and will never attract enough members to represent a statistically large proportion of the population (or even a large proportion of the intended membership). However, it is very easy for journalists to use these figures as 'evidence' that a cause has attracted a lot of attention. Typically, no mention is ever given to whether the number of members is a small or large fraction of the target membership.

In the past, when you wanted to register your protest over the latest issue du jour, you might write to your politician (this is something that I used to do frequently).  It was also common for campaigns to collect written signatures and present a petition to the White House or to 10 Downing St. If you live in the UK, the petition process has become digital and you can save yourself valuable ink when expressing your disgruntlement and just reach for the mouse and keyboard instead. At least with these 'e-petitions' you still have to include a name, email address, and postal address. But in the era of facebook, we can be content in not having to even type a single character. Click once to see the group, and click once more to join it. The facebook form of digital democracy means you can vote with one hand, leaving the other hand free to finish off your Big Mac.

My point is that joining a facebook group to register your support for or against an issue takes so little effort as to be almost meaningless. I wonder how many people have notionally joined worthy causes whilst simultaneously viewing the latest 'Hitler-in-bunker Downfall parody video' or updating their crops on FarmVille. I therefore feel that journalists should also not attach too much importance to facebook groups.

A recent article in the University of Maryland's student newspaper (The Diamondback) noted that a student protest group on facebook 'grew to nearly 1,000 members'. Nearly? It didn't even get to four digits? If we assume that nearly all students have facebook accounts, then that 'nearly 1000' figure — and I'll be generous and round up to a full 1000 — represents less than 3% of the student population of the University of Maryland. Unsurprisingly, if the journalist had written 'less than 3% of students were bothered with this issue', this would not seem to newsworthy.

Here's another student newspaper article that also cites a facebook protest group that has ... wait for it ... 61 members! This isn't just an issue with small, student newspapers. Here's an article in the New York Times (via the Associated Press) that talks about a protest group on facebook for 'military spouses'. There are nearly 1.5 million people on active duty in the American military (not including reserves). I don't know many of those are married, and I also don't know how many of the married service men and women are on facebook. However many it is, I bet it is a lot more than then 1,200 people that the New York Times article draws your attention to.

Hopefully, I have made my point. These examples are not newsworthy. I could probably name 100 friends and colleagues who would all agree with me in that they oppose kittens being kept in saucepans for extended periods of time. I don't know if kittens-in-a-saucepan is a pressing issue or not. Whatever it is, it is not news. And yet, if I were to get those same people to join a facebook group with the same name, then it suddenly has much more chance (admittedly still not much) of being cited in a news article.

In closing, I would say that there are only two facebook groups that I would be happy to join:

1 million people (but preferably a suitably large proportion of the target demographic) to stop creating facebook groups that target 1 million people

and...

1 million people (but preferably a suitably large proportion of the target demographic) to stop journalists writing about facebook groups