Is Facebook defriending on the rise? Don’t believe the Pew Facebook myth

Imagine if 10% of Facebook users defriended five of their friends on the social network each year. Both numbers static year after year. I suspect most people would describe that as defriending levels staying static on Facebook.

Take it one step further. Imagine that it isn’t 10% and five people, but each number going down – 10% and 5 people two years ago, 9% and 4 people last year, 8% and 3 people this yet. That sounds like a falling trend, doesn’t it?

But in both cases the percentage of Facebook users who have ever defriended at least one person could be going up. Even if there’s heavy overlap between the people who defriend one year and the next, if there’s not complete overlap, the total who have ever defriended someone would be going up even as the action is becoming less popular and less widespread.

Or to flip it around, knowing that a higher percentage of people have at some point in the past defriended someone doesn’t tell you whether more or less defriending is really taking place. There could even be fewer people carrying out fewer defriendings year after year but the number who have done it at least once goes up.

Why does this matter?

Because it’s a failure to understand this that has resulted in a mini-blizzard of stories in the last few days such as Facebook defriending is on the riseJudging your popularity by your number of Facebook friends seems to be falling out of fashion and Facebook friend lists shrinking to guard privacy.

All these stories are based on a Pew Study which actually says something rather different from the way nearly all of its coverage puts it. All the study shows is that the number of Facebook users who have defriended at least one person once has gone up. Yet as my examples above show, that could be quite compatible with the level of defriending falling year on year.

Is it? We don’t know, because the other evidence that would tell us if it’s true hasn’t been provided.

But that hasn’t stopped the stories which a bit of simple maths shows are drawing a conclusion the evidence does not support.

Leave a Reply

Your email address will not be published. Required fields are marked *

All comments and data you submit with them will be handled in line with the privacy and moderation policies.