#samaritansradar

[isufferfrom]

What do you think about an app that scans your Tweets for suicidal behavior and lets your followers know about it? Would you be more likely to hide your feelings because you might be "found out" or would you be more open, knowing that people would be there to help?

[ovalsilver]

creeped out by the phrase "turned my social net into a safety net"

samaritans radar app seems to be yet another good idea gone weird

http://www.dailydot.com/opinion/samaritans-radar-online-privacy/

[queercrip]

Error in now deleted Samaritans Radar whitelist post:

In my last post, which has now been deleted (it wasn't up more than 5 minutes), I referred to the Samaritans failure to whitelist a user 6 hours after their request as a "whitelist failure". I apparently overlooked the part of the Samaritans Radar website that mentions that it may take up to 72 hours to be whitelisted after the request. 

Sorry about that, and thanks to my twitter followers who corrected me.

[tjathurman]

The Samaritans, who are generally a wonderful organisation, have accidentally produced a tool to enable bullying. But they have not yet realised what they have done, and are still protesting that all it does is use public information.

A few years ago, someone wrote an app that combined Foursquare and Facebook data to alert you when single young female users checked in to coffee shops or bars nearby. Then there was the app that alerted you when people whose home address was known and nearby checked in somewhere a day's journey away. Both of these used public information, and both of them could have been done by anyone who was handy with scripting, but both of them were a bad idea for the same reason the Samaritans Radar is a bad idea. [Yes, I know: I can talk, I wrote Joule.]

[queercrip]

Testing Samaritans Radar - part 1: What I'm Doing to Test This Horrible App

TRIGGER WARNING: phrases related to depression, suicide, and self injury were used for tests. Please make sure you're safe to read this. ♥

I'm conducting some tests on the Samaritans new Radar app on twitter. If you're not familiar with what that is, the gist of it is that if one of your followers uses Radar, they'll get notifications if you tweet certain phrases deemed "worrisome" by the Samaritans (more details here). If you don't know why it's a terrible idea, take a look at this excellent summary by Adrian Short. But basically, here are some of the problems: 

People struggling with mental illness (or just saying "worrisome" things) may not want or need help.

Many people who use Twitter for support now feel unsafe doing so. (Note: This problem is a little better now that you can opt out – you couldn't at first. But it doesn't help if someone doesn't know about the app & is contacted unexpectedly, causing anxiety about reaching out in the future. They still won't know about it unless the person contacting them tells them about it.)

Radar users aren't trained, so even well intentioned users can do a lot of harm.

You don't choose your twitter followers, so you can't choose who's monitoring you.

Abusers can use it to see when victims are most vulnerable.

What people with mental illness do and say is NOT public property.

CONSENT. CONSENTCONSENTCONSENT.

You don't know who's using it or who gets notifications about your tweets. You are NOT notified when a follower installs the app OR when they receive a notification about you. 

You can opt out of Samaritans Radar by direct messaging @Samaritans, but that doesn't help people who don't know about the app or don't recognise the potential for abuse (which the Samaritans adamantly refuse to acknowledge). (EDIT: Most people are unable to message @Samaritans without following them. To opt out, email [email protected])

WHY I'M DOING THESE TESTS

Knowing more about what it actually does means we can more accurately figure out things like how potentially dangerous it is and how well it works.

Some proponents of Samaritans Radar say "You don't even know how it works!" or "Have you even tried it?" to dismiss the concerns of people with mental illness and people vulnerable to abuse.

I think it's important to make sure that opting out (they call it being whitelisted) actually works consistently. 

HOW THE TESTS WORK

I currently have three twitter accounts involved in this: my regular twitter, @BathysphereHat, and two new ones, @InspectorRabbit and @RadarTestRabbit. I've installed the Radar app on @InspectorRabbit and followed my other two accounts. As @RadarTestRabbit and @Bathyspherehat, I tweet test phrases to see if the app notifies me via the email address I gave them as @InspectorRabbit. That way, I can find out what phrases Radar flags as "worrisome". 

As @InspectorRabbit I've also followed a few users who have volunteered and consented to be monitored using Radar. I do NOT follow anyone on that account without their consent. 

There are also a few other lovely, lovely people who've made dummy accounts similar to @InspectorRabbit to help. 

I begin all test tweets with *TEST TWEET*, but there were still some issues with people being concerned and confused at first, so I now also add a bit at the end to hopefully make it clear that I'm ok. Unfortunately, I can't start test tweets with "TW:suicide", because the word "suicide" trips the app.

Test tweets look like this:  

I'm also helping people who've opted out to test whether they've actually been whitelisted. If you consent to me temporarily following you as @InspectorRabbit, I'll give you a phrase confirmed to trigger Radar notifications. If I don't get a notification within an hour, I'll tell you you've successfully opted out & unfollow you. If I do get a notification, I'll let you know that you haven't been whitelisted, and then I'll have strong words with @Samaritans.

So that's how the testing works! I'll post about what I've found out later.

Tweet at @InspectorRabbit about it if you've opted out and want to know if it worked.