In which Microsoft stands in a cesspool and declares it fit for swimming

February 6, 2019 12:35 PM

People around the world report increased civility online, new Microsoft research shows. The survey [Powerpoint] questioned teens and adults in 22 countries about their exposure to online risks across four categories: reputational, behavioral, sexual and personal/intrusive. There were tiny improvements worldwide in perceived online civility, with US showed the biggest positive change.

Microsoft sees a tiny improvement in the cesspool we call the internet: Before you break out the tea and crumpets to celebrate the return of politeness, there are some sore points. For one, the consequences of dealing with incivility were more prevalent, with more people (three to four points in each category) becoming distrustful, stressed, sleepless and withdrawn when situations turned sour. People were also less likely to think before replying, treat others with dignity or stick up for victims. [...] On top of this, we'd point out that Microsoft is judging internet civility by its own criteria -- we're sure many people would beg to differ.

Microsoft Study Claims the Internet Is Getting More Civil. The Internet Throws Side Eye Back: Though it seems there’s a reason why people are feeling like the Internet is slightly more civil than it was in 2017: they’re starting to unplug more. The survey found that 27% of people—a 4% increase—said all of the negativity has prompted them to spend less time participating on social media and blogs.

posted by not_the_water (16 comments total)

It seems like you are trying to survive online.

Would you like help?
posted by Foci for Analysis at 12:45 PM on February 6, 2019 [6 favorites]

Maybe because everybody's leaving facebook?
posted by rikschell at 12:52 PM on February 6, 2019 [1 favorite]

Maybe because everybody's leaving facebook?

Quite the opposite.
posted by EmperorOozy at 12:57 PM on February 6, 2019 [1 favorite]

People around the world report increased civility online, new Microsoft research shows.

The hell you say...
posted by Fizz at 12:59 PM on February 6, 2019

With only 500 surveyed in each country across an age range of 13-74 and across countries as diverse as Colombia and Vietnam and Hungary, it's very difficult to draw meaningful conclusions. And when all the answers are fed into a "Digital Civility Index" of Microsoft's own devising which attempts to formalise a subjective experience (how does it feel to be on the internet?), it's little surprise that the common reaction is a resounding "eh?"
posted by humuhumu at 1:19 PM on February 6, 2019 [3 favorites]

seems like the main takeaway from this study is perennial good advice: "log off", "never tweet", etc.
posted by vogon_poet at 1:32 PM on February 6, 2019 [1 favorite]

> Maybe because everybody's leaving facebook?

Quite the opposite.

Anecdotal, but the amount of activity from my friends on Facebook has dropped dramatically since 2017. A bunch have permanently deleted their accounts explicitly because of the current politics around social media and its influence on elections as well as social divisions. Many others have kept their accounts but don’t actively use them anymore. In general i see fewer postings/comments from the people who remain, compared to their level of activity during the Obama years.

It’s entirely possible that Facebook can truthfully say their overall numbers are growing but also that they are seeing (in some demographic groups, anyway) much less activity from existing FB members as well as an increase in account deletion compared to what they would have projected from past statistics.
posted by D.C. at 1:32 PM on February 6, 2019 [4 favorites]

I'd say this seems anecdotally true. When I first started using the internet, there wasn't even vocabulary for all sorts of toxic behaviour online. Toxic wasn't even a word used that way then. If you got offended or hurt by something, it would have been rare for someone to comfort or get mad about it other than telling you to get over it and get off the internet if it bugs you so much. There are now whole communities built around fostering safe spaces or banding together against hateful mob actions and such.

There's still tons of nastiness online, sure, but there's tons of nastiness in humanity.
posted by GoblinHoney at 2:17 PM on February 6, 2019 [5 favorites]

I get this impression that the research in question would consider “Nice store you’ve got here. It’d be a shame if anything happened to it” to be “polite”
posted by DoctorFedora at 2:24 PM on February 6, 2019 [2 favorites]

Somebody get that 'this is fine' cartoon and label the dog 'Microsoft' and maybe have it saying 'this is increasingly civil'. I can't be bothered.
posted by Sing Or Swim at 2:34 PM on February 6, 2019 [4 favorites]

Minor pet peeve, required sample size depends on power and effect size not the the demographic population of the country in question. It's counter-intuitive but it's a common fallacy. If your claim is "Comet, it makes you vomit," you can probably get away with a very small sample size because almost all of your experimental group will need hospitalization. But you'll also need a very good lawyer to get away.

I don't doubt their claims to statistical significance at 95% given their sample size and comparing the same data collection between two different years. Is that practically significant? I found a 99% significance that involved an imperceptible difference between groups. That result was absolutely useless except as a case study in the problems of crunching numbers from large automated samples.

But the bigger bias here from skimming the PowerPoint (which still isn't sufficient to evaluate the methodology or the results) is that the Microsoft blog post is cherry picking some of the better results in order to grab headlines, and Fortune and Egadget parroted what is basically a Microsoft publicity release. This, unfortunately, is the ugly state of science reporting.
posted by GenderNullPointerException at 2:44 PM on February 6, 2019 [2 favorites]

"Journalists" whose only source is a corporate press release probably languish in a layer of hell just above those whose only source is Twitter.

At the center are the "president tweeted something" media.
posted by GenderNullPointerException at 3:07 PM on February 6, 2019 [3 favorites]

Increased civility = people finally figured out how to use those "mute" and "block" options
posted by talos at 4:11 PM on February 6, 2019 [3 favorites]

Just so you know, "Microsoft Research" is a quasi-academic group, like Bell Labs was to AT&T. They have a lot of very qualified scholars and researchers there. I don't really understand this work, but I don't think it's corporate crap.
posted by Harvey Kilobit at 4:26 PM on February 6, 2019 [1 favorite]

I don't think the research is crap, but part of the business of research are marketing releases and blogs written to get the sponsors in the news. And I've encountered plenty of cases where key findings were creatively filtered and interpreted to reduce ambiguity and support an easy story.

While Microsoft does great research on online social networks, you wouldn't know it from the the marketing material linked in this post.
posted by GenderNullPointerException at 5:09 PM on February 6, 2019 [3 favorites]

Somebody get that 'this is fine' cartoon

Ask and you shall receive.
posted by EmperorOozy at 8:45 PM on February 6, 2019 [7 favorites]