Manipulating Beliefs and Behaviour via Social Media

Manipulating beliefs and behaviour via social media

Software	None
Covers	Ethics, psychology, social impact of the internet, social impact of technology
Skills Required	Interest in ethics and moral philosophy, interest in social impact of computers, interest in psychology, ideally some understanding of statistics
Challenge	Conceptual ???? Technical ?? Programming

Brief Description

How information is presented on social media systems such as Facebook, and what information is selected for presentation, as well as how users then pass it on, might have a significant influence on people's beliefs, attitudes and behaviour. What is it acceptable to control or alter, how, and when, and why?

This information can be manipulated for purposes of scientific research. In 2014 Facebook published a study finding evidence for large scale 'emotional contagion' produced by users transmitting positive or negative emotional messages to each other by Facebook, influenced by the emotional content of the Facebook newsfeed. This got them - and Cornell University - a lot of adverse publicity for violating ethical standards expected for research. Cornell was forced to make a public statement that since its researchers used data previously gathered by Facebook, they hadn't needed to take their study to an ethics review.

This information might also be manipulated for commercial gain, or to persuade people to hold particular beliefs or attitudes, by selective filtering of information, choice of descriptive words or framing contexts, and by outright distortion or lying - all familiar tactics used by businesses and political groups with other means of communication. For instance, American political advertising frequently involves lying by implication without making individual statements that are demonstrably false - if you're well-informed about the facts you're not being told, you can see that you're being bamboozled, but all the voters who aren't... And the politicians being slandered can't scream "That's a lie!" - they can only hope that the voters will listen to a complicated explanation.

Systematic distortion and manipulation doesn't need to be under direct human control or even intended. In February 2018, The Guardian published an article reporting on how YouTube's Up Next algorithm directs attention to extremist views, fake news and hate speech, systematically biasing the consumption of political content in pro-Brexit, pro-Trump, racist and islamophobic directions, just because the extreme content is effective clickbait. There is a serious possibility that this has influenced close elections.

Variants

A project on how social media applications such as Facebook can be used to manipulate users' behaviour could have a number of different slants.

Ethics of research. How far, and in what circumstances, is it justifiable to bias the information transmitted through social media? Might the value of understanding how social media influences moods, attitudes and beliefs justify some experimentation without consent or other practices that an ethics committee might look twice at? What value might this have? What is it possible to do, and how? Are there ways of neutralizing the effects of biasing experiments, or can two influences never add up to zero? Can we investigate what biases in information presentation occur naturally?
Ethics of information presentation of for commercial or political gain. Commercial organizations are in the business of manipulating beliefs and attitudes: it's called advertising and it's usually regarded as legitimate if it is clearly labelled and doesn't make false claims. But sometimes commercial organizations want to influence the presentation of other information. Similarly political and religious organizations and pressure groups want to influence what people think, and are utterly convinced that they are morally right in doing so. What are appropriate moral limits on how information transmitted through social media is filtered or biased? Are there any means to counteract or control this that wouldn't do more harm than good, whether political, legal, or done collectively by social media communities.
Living in one world. One effect of the universal use of the world wide web and social media like Facebook and Twitter, is that for better or worse people have a wide choice of sources of information. This makes government censorship and control of news reporting much harder, which is why many governments want to control what their citizens can access. It also enables communities to develop that become increasingly isolated from their neighbours as their understanding of the world comes to depend on a self-reinforcing network of communication transmitting very different ideas and beliefs. In democratic western countries we see groups of political or religious extremists who aren't living on the same planet as the rest of us, in that their beliefs about what is happening in the world and what other people do and think is disconnected from reality. Many of these people might abhor political violence but contribute to the development of a collective worldview that is conducive to justifying terrorism. How is information framed and filtered as it passes through social media, and what influence does it have? Is there anything that can be done about the phenomenon of social fragmentation? If so, is doing it morally justified? Would it do more harm than good ? Is there any way of judging what is fair and what is distortion?
Technology and ethics of controlling content advertising. How much of a social and political problem are automatic recommendation systems directing people's attention to videos and other content related to what they've looked at before, that bias the choice of recommendations in particular directions? What can be done to counteract this, technologically, legally, socially, and all of these in combination? What is technologically, legally and socially practical to do? What undesirable side effects would such action have? What technological or legal regulatory action would be ethically acceptable? Where are the ethical complications?

Resources

A short paper produced by the controversial 2013-2014 Facebook study:

Kramer, A.D.I., Guilloy, J.E. & Hancock, J.T. (2014) Experimental evidence of massive-scale emotional contagion through social networks. Proceedings of the National Academy of Sciences, 111, 8788-8790.

Lying by Sissela Bok would be a good place to start considering the morality of deception.

Lewis, P. (2018) 'Fiction is outperforming reality': how YouTube's algorithm distorts truth The Guardian, 2 February 2018.

Back to