Just my 2 cents, but I'm going to wager that facebook didn't sufficiently anonymize the data. I'm basing this on the fact that facebook merely skewed the existing data.
I wish government officials would at least get with the 1990's. I'm guessing this was little different than switching between the 'hot' and 'controversial' tabs on a reddit page. No data was changed or modified, just the way it was displayed.
The technical complexity is at the level of AB testing, but the effect is emotional manipulation.
This is ethically murky as people don't sign up to be emotionally manipulated by FB for $. People assume that they would be shown all the good and the bad from their friends.
Exactly. The issue here isn't the anonymization of data (people are jumping on this immediately due to the fact that Facebook is known for their disregard for privacy). Its the fact that they're deliberately manipulating emotions for science without consent. What if they inadvertently persuade someone to self-harm? Are they willing to take responsibility for it?
That seems as a bit of a slippery slope. Also, I dislike FB as much as the next guy but I can't really imagine that someone would sign up for FB without realizing that something like this feed manipulation is a real possibility.
There are laws regulating the use of human subjects in experimentation. I have faint but enough knowledge to believe that Facebook might not be in the clear here.
As a person who works in IT, and from my own personal point of view, I agree that I find it hard to imagine that someone would sign up to FB without clear awareness that feed manipulation and other shenanigans might occur[0]. But the thing is, people are doing so every day.
Most people on FB are using FB assuming it's a utility something like email, without realising that it is like email, if email had its own agenda. It's really easy to forget that most people in this world do not live/work in the echo chamber that we do. When I talk to them about the issues, sure, at best they might profess to understand or even be concerned at the time of that conversation, but by and large they won't change their behaviour and the conversation will be quickly forgotten.
Never underestimate the indifference with which people can view the inner workings of technology.
0. I see the fact that I find this hard to imagine as a form of "Expert Blindness" or "Knowledge Blindness". If someone has a better or more accepted term for this phenomenon, I'd like to hear it.
Well I understand that feed manipulation is expected. After all, everyone knows they manipulate their feeds to expose sponsors and advertising. I have no problems with the actual manipulation.
My problem is with the goal of the manipulation in this case. Its not something like "persuade people to view pages". Its (as I understand it) "make people sad" or "make people happy". As someone who personally battles with emotional episodes, could something like this have inadvertently triggered an episode for me? Can I prove it?
Some people have put forward the argument that advertising is by definition a mood influencer and we see it everywhere. However, that doesn't mean that we allow all advertising. We frown on tobacco advertising because it influences people to buy an unhealthy product. We frown on displaying unhealthy bodies because of the impact it has on our self-images.
tl;dr; I'm not saying feed manipulation is bad. I'm saying the goal of the manipulation in this case was potentially ill-conceived.
The thing is I'm sure Facebook didn't build the AB test framework for this study. We know not all your friends posts show up in your feed. They certainly have been using AB testing to improve what stories they show to users. I also bet they have been using this framework to optimize ad revenue. This causes a problem if the study was wrong because it manipulated users then wouldn't all AB testing be wrong for the same reasons.
Isn't everything in the media designed to manipulate our emotions? I mean, those commercials to sponsor children in Africa, trying to make you feel guilty, and how for just the price of your morning coffee you can keep Timmy off the streets, and eating well?
You can argue FB relies on user content, but then I can compare that to reality TV shows, and how they'll take real-life footage, and manipulate it to trigger emotions.
For example, American Idol will paint a picture of someone struggling their whole life, and how this is their one opportunity for success after being fired from McDonalds and having their cat pass away. Then just as you're feeling sorry, they light up the stage, and you rejoice. Meanwhile, they forgot to tell you that person had professional training twice a week for the past 10 years, they won two other singing competitions earlier in the year, and live in a beautiful neighborhood, but that doesn't play into the emotions they want you to feel, and that doesn't get you to watch next week.
Or a show like Survivor, and how they'll take footage, and try to fabricate relationships and drama out of thin air. This way you become angry towards one character that'll be in the finale, and are likely to tune in to ensure they don't win.
Why can't Facebook do the same? Why can't they analyze and manipulate user emotions to increase business?
Watching a show and getting emotional is something that we do of our own will.
In my lab, a PhD student had to get permission from the IRB to conduct a simple study that looked at how good people were at critical thinking. There are reasons why this is regulated.
We do at our own will? So if I'm watching a movie and everyone in the theater is in tears, it has nothing to do with the director trying to create a sense of emotion and sadness, but it was just our own choice?
I'd think Facebook has even less control, and more difficultly predicting emotions. For example, if Facebook displays a post about Jane having a bad day and losing her job, that's bad news. However, it's difficult to determine how I'll react. It might be comforting for me to know someone else is having a bad day, it might make me angry that Jane lost her job, when I lost my job at the same business the week prior, it might make me happy because Jane is always bragging about her job, and I no longer have to hear about it.
When it comes to a movie, I think there's a lot more control since you write the script and characters from start to finish. Every person in the audience has the same relationship with those characters, and knows them for their entirety. You also have fine grain control over the visuals, combined with carefully selected music. As I said earlier, all of this can lead to a room full of people leaving the theater in tears, so I don't see the difference.
Or, when you say we do at our own will, you mean we make the choice to visit the theater in the first place? That would be no different than making the choice to visit Facebook. If anything, you should be questioning every advertising campaign in existence. They're carefully crafted to evoke a certain emotion, and they work specifically because they can manipulate people. At the same time, people have no choice to view them, they're constantly exposed to these manipulations just by walking outdoors or visiting the store to buy groceries.
Going to the theater is our choice. We want our emotions to be changed when we go to a movie. In Facebook, I want the raw feed from my friends, not some emotionally filtered feed. I don't want to get into a debate on free will. Do you think laws on human experimentation should be removed?
You argument is basically a milder analog of "Humans die from all kinds of causes so let us let murderers walk free."
Right, you want the raw feed, but it's up to Facebook whether or not to provide it, and it's up to you whether or not to consume it.
This is no different than a television series like I mentioned earlier. You can argue you want the raw footage from a reality television show, and not the heavily edited version designed to manipulate your emotions, but that's not your choice.
Where does it say they provide the raw feed? Everywhere I look in the Facebook help, terms and privacy policy it mentions how they use algorithms to determine what stories appear, and how they use information provided by users to pick stories. They also mention using user information for internal testing and analysis.
They seem to be following those terms, they were choosing positive and negative stories for feeds, and then analyzing the data to see if users then posted more positive or negative posts in return.
1. '...we may make friend suggestions, pick stories for your News Feed, or suggest people to tag in photos...'
2. 'The News Feed algorithm uses several factors to determine top stories...'
3. 'How we use the information we receive... for internal operations, including troubleshooting, data analysis, testing, research and service improvement.'
The UK is free to look into the Facebook experiment and to pursue the case. That alone doesn't make Facebook guilty of anything. They didn't go to a court of law, Facebook wasn't proven guilty of breaking the UK Data Protection Act, they're just investigating whether proper precautions were taken.
Facebook should obviously be bound to the same laws as everyone else. Which human experimentation laws did they break? Users registered on Facebook and agreed to the terms of service, and how Facebook will choose which stories they view, and analyze their response.
I hope everything that manipulates people gets hunted down with the same vim & vigor.
But you don't see people so outraged during election cycles, or wondering why they bought that widget they didn't really need, etc...
It's naive to think Facebook is the only one who has done something like this. I respect that complacency is bad, but my point is: so is singling one company out when a lot of others are getting away with it.
UK data protection laws are much tighter than in the U.S. One of the key provisions is that data must only be used for the purpose for which it was gathered. While the terms allow them to use the data for research, the ICO may decide this is not the kind of research that users signed up for. The ICO may even decide that psychological research counts as medical information, which has lots of other safeguards.
22 comments
[ 2.9 ms ] story [ 62.8 ms ] threadI wish government officials would at least get with the 1990's. I'm guessing this was little different than switching between the 'hot' and 'controversial' tabs on a reddit page. No data was changed or modified, just the way it was displayed.
It was AB testing with winks and frownie faces.
This is ethically murky as people don't sign up to be emotionally manipulated by FB for $. People assume that they would be shown all the good and the bad from their friends.
http://en.wikipedia.org/wiki/Human_subject_research_legislat...
Most people on FB are using FB assuming it's a utility something like email, without realising that it is like email, if email had its own agenda. It's really easy to forget that most people in this world do not live/work in the echo chamber that we do. When I talk to them about the issues, sure, at best they might profess to understand or even be concerned at the time of that conversation, but by and large they won't change their behaviour and the conversation will be quickly forgotten.
Never underestimate the indifference with which people can view the inner workings of technology.
0. I see the fact that I find this hard to imagine as a form of "Expert Blindness" or "Knowledge Blindness". If someone has a better or more accepted term for this phenomenon, I'd like to hear it.
My problem is with the goal of the manipulation in this case. Its not something like "persuade people to view pages". Its (as I understand it) "make people sad" or "make people happy". As someone who personally battles with emotional episodes, could something like this have inadvertently triggered an episode for me? Can I prove it?
Some people have put forward the argument that advertising is by definition a mood influencer and we see it everywhere. However, that doesn't mean that we allow all advertising. We frown on tobacco advertising because it influences people to buy an unhealthy product. We frown on displaying unhealthy bodies because of the impact it has on our self-images.
tl;dr; I'm not saying feed manipulation is bad. I'm saying the goal of the manipulation in this case was potentially ill-conceived.
You can argue FB relies on user content, but then I can compare that to reality TV shows, and how they'll take real-life footage, and manipulate it to trigger emotions.
For example, American Idol will paint a picture of someone struggling their whole life, and how this is their one opportunity for success after being fired from McDonalds and having their cat pass away. Then just as you're feeling sorry, they light up the stage, and you rejoice. Meanwhile, they forgot to tell you that person had professional training twice a week for the past 10 years, they won two other singing competitions earlier in the year, and live in a beautiful neighborhood, but that doesn't play into the emotions they want you to feel, and that doesn't get you to watch next week.
Or a show like Survivor, and how they'll take footage, and try to fabricate relationships and drama out of thin air. This way you become angry towards one character that'll be in the finale, and are likely to tune in to ensure they don't win.
Why can't Facebook do the same? Why can't they analyze and manipulate user emotions to increase business?
In my lab, a PhD student had to get permission from the IRB to conduct a simple study that looked at how good people were at critical thinking. There are reasons why this is regulated.
http://en.wikipedia.org/wiki/Institutional_review_board
I'd think Facebook has even less control, and more difficultly predicting emotions. For example, if Facebook displays a post about Jane having a bad day and losing her job, that's bad news. However, it's difficult to determine how I'll react. It might be comforting for me to know someone else is having a bad day, it might make me angry that Jane lost her job, when I lost my job at the same business the week prior, it might make me happy because Jane is always bragging about her job, and I no longer have to hear about it.
When it comes to a movie, I think there's a lot more control since you write the script and characters from start to finish. Every person in the audience has the same relationship with those characters, and knows them for their entirety. You also have fine grain control over the visuals, combined with carefully selected music. As I said earlier, all of this can lead to a room full of people leaving the theater in tears, so I don't see the difference.
Or, when you say we do at our own will, you mean we make the choice to visit the theater in the first place? That would be no different than making the choice to visit Facebook. If anything, you should be questioning every advertising campaign in existence. They're carefully crafted to evoke a certain emotion, and they work specifically because they can manipulate people. At the same time, people have no choice to view them, they're constantly exposed to these manipulations just by walking outdoors or visiting the store to buy groceries.
You argument is basically a milder analog of "Humans die from all kinds of causes so let us let murderers walk free."
This is no different than a television series like I mentioned earlier. You can argue you want the raw footage from a reality television show, and not the heavily edited version designed to manipulate your emotions, but that's not your choice.
They seem to be following those terms, they were choosing positive and negative stories for feeds, and then analyzing the data to see if users then posted more positive or negative posts in return.
1. '...we may make friend suggestions, pick stories for your News Feed, or suggest people to tag in photos...'
2. 'The News Feed algorithm uses several factors to determine top stories...'
3. 'How we use the information we receive... for internal operations, including troubleshooting, data analysis, testing, research and service improvement.'
You did not answer my other question. Here is a more direct question.
Should Facebook be exempt from laws on human experimentation?
Facebook should obviously be bound to the same laws as everyone else. Which human experimentation laws did they break? Users registered on Facebook and agreed to the terms of service, and how Facebook will choose which stories they view, and analyze their response.
But you don't see people so outraged during election cycles, or wondering why they bought that widget they didn't really need, etc...
It's naive to think Facebook is the only one who has done something like this. I respect that complacency is bad, but my point is: so is singling one company out when a lot of others are getting away with it.