The pocket-sized Big Data made up of everything you do, feel and decide.
By Renata Neves
Source: Freepik
It sounds like a conspiracy theory, but it is real: all our movements online, or the lack of them, are being monitored and analysed. This makes it easier to predict patterns of behaviour among users and to determine the kinds of content that steer our decisions.
These are the behavioural data that are stored, sold or shared for marketing strategies, political campaigns and other activities that need consumers to take part.
Digital platform algorithms are also being used, however, to capture and analyse people’s psychic and emotional data.
In other words, it is now possible to use technology to predict when a specific group of users is sad, happy, tired, vulnerable and so on. That is the moment when marketing or politics moves in with emotional content, exclusive to that group, securing the right behaviour at the right time.
At the end of the story, we have a huge mass of people being steered towards making political or commercial decisions.
All of this works like a giant laboratory in which various social experiments are run by companies or politicians on the platforms, taking advantage of abusive terms and conditions and assembling inferences about users, even without any certainty that the algorithm is always right.
And when the technology gets it wrong, in these cases, the effects are large. As happened in 2018, when Donald Trump won the presidential election.
It is a lot of information to take in and accept, but we are going to explain slowly how the psychic economy works and how it is being used in your life. So, shall we?
Understanding the Psychic Economy of algorithms
The concept of the Psychic Economy was developed by the researcher Fernanda Bruno while studying the way digital platforms capture behavioural data from their users.
According to her, the psychic economy of algorithms is the investment in processes of capture and analysis, as well as the use of psychic and emotional information extracted through monitoring.
In other words, the algorithm is built to determine a percentage for some psychological characteristic that you display through your actions in an app.
For example, if you listened to type ‘x’ of music, skipped quantity ‘y’ of others and took an interest in type ‘z’ of recommended albums, the technology is then able to draw your profile in real time and store your behavioural pattern.
Yet all of this happens without your knowing and, worse, without your being aware of why.
The psychic economy arose within today’s data capitalism so as to make it easier to control the environment in which people are most influenced.
The point of having a base of psychological information is to generate profit, even if the process is ethically suspect.
Your psychic and emotional data are used to facilitate sales, target political campaigns, influence your actions in moments of vulnerability and create patterns of behaviour across various groups of people.
In this way, digital platforms and whoever is behind them will be ready to send the best content to redirect your decisions.
The technologies that emerge from the psychic economy undertake to predict the user, as though the person behind the phone were irrational and impulsive.
In short, the algorithm gathers various patterns of behaviour you display in apps, draws a profile and manages to make predictions about your decisions, enough to guide marketing strategies and electoral campaigns.
In an interview, the researcher Fernanda states that “the future is being hijacked”, precisely because all your options and the diversity of possibilities are narrowed to a single direction, a single choice.
She also explains that digital data and psychological information make up three layers: the economic or market layer (aimed at profit from the exchange, sale and use of the information), the emotional management and control layer (content targeted to provoke specific emotions), and the epistemological layer (the production of knowledge about a subject or a group of subjects).
This means that the positive side of the process lies in the acquisition of knowledge. Science thereby expands and there is more development in technological innovation.
However, a psychological study of this kind requires millions of experiments that are being run without a scientific basis, on millions of people and, still less, without users’ consent. All with the aim of generating more influence towards particular commercial or political directions.
Do you realise that there are many tests predicting your digital actions in the hands of large corporations, politicians and platforms?
The host of the Tecnopolítica podcast, Sérgio Amadeu, recalls that every experiment has its margin of error. So if this is being applied at large scale, the percentage of error also represents a huge number of the general population.
The basis of the prediction and control strategy
It is important to be clear about the theoretical basis of the method for capturing and assessing emotions.
You can read Fernanda Bruno’s full article to understand the references and the theorists cited, by clicking here.
A model of profile design using demographic analysis was already widely used, capturing data made available by people, such as gender, age, where they live, where they work and so on.
The new strategic model, however, carries out data analysis based on a psychometric profile, personal data and digital relationships.
One of the differences between these two models is that the first does not necessarily mean being entirely sincere, which is why it becomes necessary to analyse users’ natural actions in order to build the predictions.
The logic behind the strategy is: determine patterns of behaviour, define a psychological profile, compare the patterns with other users, define groups of profiles and observe the interaction between them.
In this sense, the machine becomes able to predict what the next behaviours will be in the face of a piece of content and so marketing and campaign strategies can alter the environment in their favour.
Changing the environment to alter or shape a behaviour and creating content that conditions people are characteristics that originate in behaviourist theory. And this is where we want to get to.
The premise of behaviourism, as Anna Bentes states, is the prediction and control of behaviour, just as we see in the psychic economy.
The theory emerged in the nineteenth century with the psychologist John B. Watson and sought to be developed as an objective, measurable science; it was therefore necessary to run experiments that could be described and measured. In fact, every scientific study needs a basis that can be proven with observations or numbers, but how would it be possible to predict human behaviour? That is what we need to understand.
Watson defended Methodological Behaviourism through classical conditioning. That is, for him, associating elements with stimuli could even shape or adjust a child to take up any kind of profession.
One of the experiments carried out by the psychologist was to develop fear in a newborn by associating a rat, which caused no reaction in the child, with a loud, alarming noise.
By the end of the tests, the baby began to have episodes of distress and fear on seeing the animal or things resembling it, even without the sound being present.
Skinner, for his part, sought to demonstrate Radical Behaviourism through operant conditioning. He explains that the use of positive or negative reinforcement and punishment produces consequences that adjust behaviour.
For example, in his test “the Skinner Box”, a rat repeated its behaviour whenever it was rewarded. But the frequency of the behaviour fell when it received punishments.
Even so, both theories treat the object of study as an irrational being, disregarding its thoughts, feelings and emotions. In this sense, it is possible to find a relationship between the theory of Behaviourism and the practice of the Psychic Economy.
The problem is that experiments with psychological data and the use of content to influence decisions are tests being applied in countless ways and on thousands of people.
How does data capture happen?
You can think of everything: likes, searches, emails, deleted texts, photos, favourite songs, skipped songs, locations, travel speed, typing speed, page views, reading time, spelling mistakes, clicks, purchases, and other patterns of interaction.
Beyond that, there is investment in trying to extract users’ emotions, personality, all their vulnerabilities and inclinations.
Through the architecture of the platforms, reading an emoji makes it easier to identify an emotion translated into programming language, for example. That is without mentioning studies on facial reading and the improper use of biometrics.
All these smaller data (small data) are what make up big data. It stores the analysis of our relationships with other accounts and profiles, revealing to the technology of the psychic economy the patterns of behaviour.
Indeed, the central issue in analysing psychic and emotional data is that it is not interested in the characteristics of a single individual, but in how they react and relate to one another among other personalities.
In this way, knowing isolated pieces of information, such as the number of likes from any given user, may be rather useless. On the other hand, knowing that a specific type of profile has the habit of leaving a “love” reaction on colourful posts from account “x” may be valuable information.
When small data is collected, the profiling of each user is done on the basis of techniques from Behavioural Economics. Studies from behaviourism, cognitive psychology, evolutionary psychology and neuropsychology thus become the main tools for indicating the probability of each behaviour.
It is worth stressing the word “probability”, because a machine is not able to get the neuropsychic type of millions of people right every time. Even so, the strategy attracts organisations because the technology promises to predict behaviour with precision and effectiveness.
A person comes to be associated with one or more types of profile and this will completely shape the content-delivery actions that follow.
A sequence of content tests is applied to similar users, looking for more patterns and trying hard to steer the group’s decisions towards agreeing or disagreeing.
In practice, it would be like creating several adverts changing small pieces of information, colours or captions, to work out which perform best and which are most accepted. In this sense, there are no problems with errors, because they are merely learning material for the technology to predict better afterwards.
This strategy of producing versions with minimal differences of detail was one of the approaches used in Trump’s campaign in 2018. Thousands of adverts were applied to different market segments and microtargeting. We will go into detail about this in the next section.
To strengthen the capture base, monitoring has to happen in real time, assessing how each piece of content generates stimuli. So the longer a user stays on the platform and interacts, the better it is for the analysis.
This model of data capture has to reinvent itself by thinking about how to hold attention and get more engagement. At the same time, it has to compete for the relevance and the value a user would give to a digital platform.
The idea is to transform design, automation and user experience enough for using the platform to become addictive and turn into a habit.
Cambridge Analytica’s psychometric database
In 2018, the British company Cambridge Analytica, which ran Donald Trump’s political campaign, was at the centre, alongside Facebook, of one of the biggest controversies seen over the use of data and psychometrics.
Years earlier, colleagues Michal Kosinski and David Stillwell carried out a study by launching the MyPersonality personality test inside Facebook. The questions were psychometric in nature, assessed each participant’s profile type on the basis of the Big Five, and requested access to Facebook profile data.
What was meant to be a local experiment ended up involving millions of people.
Kosinski realised that it was possible to combine psychometric data with Facebook behaviour to make almost perfect inferences. The researcher showed that with an average of 68 likes per user it was possible to predict skin colour, sexual orientation, political affiliation, intelligence, religious identity, use of legal and illegal drugs and even divorce among family members. By the end of the research, his mechanism was able to assess people with just 10 likes.
In general, Kosinski’s methodology allowed the Big Five to be assigned to anyone online or offline who authorised access to their data. After all, since smartphones are tools loaded with secrets and thoughts, the researcher only needed to be a good analyst of psychometrics.
Even so, Kosinski published his study, turned down job offers and did not sell his MyPersonality database, despite being asked.
In 2015, Cambridge Analytica was hired to run the online campaign for Brexit and announced the work through innovative marketing with big data. In other words, the company used Kosinski’s very methodology, but for political ends.
Now the app that asked for access to Facebook data in exchange for a personality test was the well-known This Your Digital Life.
Given Brexit’s apparent success, the company was called in for its next job: Donald Trump’s electoral campaign, with the same marketing strategy based on psychometric data.
US voters were divided into thirty-two personalities, analysed and turned into targets for highly segmented adverts. On the basis of big data, it was possible to approach each user in a different way.
For example, for an audience with intense, emotional reactions, it was enough to present guns as a way of protecting oneself from robbery. For an audience concerned about family, it was better to show guns as a way of protecting children.
And so Cambridge Analytica managed to send different messages to each type of voter. It could even present Hillary Clinton in the worst possible ways to every type of audience identified.
Thousands of adverts were fired off to test the stimuli, Trump’s teams went to people’s homes with speeches tailored to each personality and every reaction became data to complete the database.
Casa Hacker, in partnership with Tactical Tech, produced a video with more detail about how the electoral strategy worked, which you can watch by clicking here.
And although Cambridge Analytica never accepted or spoke about its possible responsibility during the electoral race, the researcher Fernanda Bruno is right to state that experiments like this may not interfere with personality, but they will have consequences for behaviour.
The techniques, the studies, or the strategies may also involve attempts at emotional contagion (firing off a run of content that causes an emotion), real-time monitoring of the feed, buying and selling personal data, authorisation of access in the terms of a digital platform, and so on.
We have also picked out for you this video with an excerpt on the power of big data presented by Cambridge Analytica. Click here to watch it in English.
Political or commercial influence?
In this “world-laboratory”, behaviourist-based experiments are applied directly in people’s everyday lives.
That is, the theory claiming it is possible to alter the environment in order to predict and control behaviour can now count on technology to be tested and used without ethical awareness.
We have presented a real case in politics, but the Psychic Economy is also here to create market advantages.
Through studies in psychology, it is possible to find out that people with bipolar disorder may go through a phase in which there is a greater tendency towards consumption. By analysing behaviour and monitoring users’ activity, it is not hard to identify a group of bipolar people and, above all, to predict the right moment when they are most likely to make expensive purchases impulsively.
When an organisation has that information in hand, along with data to build the probabilities of that prediction, the result may be the implementation of a marketing strategy aimed at profiles with psychological disorders.
In the same way, big data can also predict when type “x” of person is feeling lower self-esteem, afraid of the future, interested in a new hobby, attentive to a cause, and various other situations.
In the end, everything will be converted into products and services to be sold at the right time and to the right audience.
On the other side of big data, we have an audience that, curiously, knows a little about the data capture situation or can build awareness of the subject through research.
Yet this same population is used to accepting the “terms and conditions” of any set of rules, just to use digital platforms. Moreover, apps promising easy money, access to pirated content and undeveloped games start to appear more often in the lists of popular downloads.
These days, people want to be seen and to feel they belong in online relationships. They carry on feeding all kinds of databases, not least because nobody can, or wants to, disconnect.
That is why knowing the personal data protection policies and understanding how attentive brands are when requesting authorisations is essential for a more conscious digital life.
The Lei Geral de Proteção de Dados (LGPD) came about to guarantee more privacy and security for the user, but it always needs improvements and updates as new platforms emerge and make space in the market.
Beyond that, you can assess the way you have been using the internet and how constantly you deposit your data online. To that end, the Digital Detox method has become a valid practice for those looking to reduce the time spent in online life in order to enjoy real life.
It is always important to remember that, despite the data capture strategies, the internet is part of our everyday lives and plays fundamental roles for health, education and innovation. Even so, we do not need to be hostage to virtual tools all the time.
Reflect on your consumption and on how your life is being affected. You can set goals and times to reduce the time in apps little by little or, if you find it interesting, try going the weekend without looking at your phone at least once a month.
We are sure you are able to review your priorities and achieve a healthier digital life!
Help us with a donation
You have read an article produced by Casa Hacker, a non-profit initiative dedicated to fighting social inequality through technology education.
Together, we can build better-informed communities on the path to development.
References
- BRUNO, Fernanda. A economia psíquica dos algoritmos: quando o laboratório é o mundo. Nexo Jornal, 2018. Available at: .