Ben Williamson
There is currently growing concern about the data mining of children, and particularly with the collection and calculation of children’s educational data through new kinds of learning analytics, real-time assessment and data visualization techniques. These techniques construct the child as a ‘data double’ known from digital traces of their activities. In this set of sketches for further research, I want to focus the ways in which the child, understood biologically, psychologically and neurologically, gets encoded through digital data practices that then fold back into new bio-, psycho-, and neuro-pedagogies of bodily, emotional and brain improvement. This means seeking to ‘anatomize’ the data-child, cutting through the digital skins, data membranes and algorithmic tissues that increasingly constitute how children are seen, known and acted upon.
Image: Stephen Hampshire
The focus is on:
- how the biological and physiological body of the child gets encoded as biophysical data that can be sensed by health tracking devices in physical education, and from there translated into ‘bio-pedagogies’ of self-quantification and bodily optimization;
- how the psychologically affective experiences of the child get encoded in ’emotional learning analytics’ through facial recognition algorithms and galvanic skin sensing devices, and converted into ‘psycho-pedagogies’ of emotional maximization;
- how the neurological plasticity of the child’s brain and cognitive functioning gets encoded into ‘learning algorithms’ and from there built in to new digitized ‘neuro-pedagogies’ of brain empowerment using cognitive computing applications.
Analyzing these developments requires a bringing-together of work around digital data with work around the bio-, psy- and neuro-sciences. It also situates the ‘educational’ development of new bio-, psycho- and neuro-pedagogies in the wider code/space of contemporary digitized environments where knowing, addressing and shaping the bodily/biological, emotional/psychological, and cognitive/neurological comportment of individuals through programmable technologies and digital data–and the data doubles they produce–is becoming increasingly central to techniques of governing.
The key point here is that the child is increasingly being traced, monitored and known through particular kinds of ‘data practices’ that are designed to record biological, psychological and neurological signals from their bodies, emotions and brains, and then acted upon by new kinds of pedagogic devices that are consequential to their subsequent bodily, emotional and embrained existence.
Body-beta: Biopedagogies of bodily optimization
Recently, there has been a growth of interest in the use of health-tracking and physical activity monitoring devices in physical education (PE) in schools. In line with the technologies of the ‘quantified self’ and ‘personal analytics’ that people use to keep track of their physical exertion, calorific intake, and much more, these ‘self-mediation interfaces with health’ as Emma Rich and Andy Miah describe them, are now ‘inextricable from the manner in which people learn about health’.
The physical education schools market is an obvious target for health-tracking products. A key emerging area is health-tracking devices and apps for students that are designed to encourage healthy lifestyles, aid dietary planning and encourage physical activity. Popular features of health-related apps for children include the concept of caring for virtual creatures by fulfilling their dietary and fitness needs, often combined with various gaming and competition elements and online social media environments. For example, Zamzee, the ‘game that gets kids moving’, combines state of the art accelerometry technologies, game design and ‘motivation science.’ It consists of a wearable ‘meter’ device to ‘measure the intensity and duration of physical activity; an online ‘motivational website’ featuring challenges and lesson plans; and sophisticated ‘group analytics’ to enable educators and school administrators to ‘track individual and group progress with real-time data.’ The strapline for the product is ‘Motivate. Measure. Manage’.
Another hybrid health gaming/social media platform for children is Sqord, ‘your online world, powered by real world play’, which consists of a wearable data logger, an online social media environment and a personalizable onscreen avatar called a PowerMe. Sqord is marketed as ‘one part social media, one part game platform, and one part fitness tracker.’ Sqord users can compete with one another on an online leaderboard through everyday physical challenges, as measured by their activity trackers, and are able to win rewards for completion of goals, which can be used to purchase upgrades and personalized features. Sqord also provides an administrative reporting tool for educators to access metrics on the physical activity levels and participation of each child player. Likewise, Polar Active is a wristwatch device that twins with an online environment to allow teachers to view, analyze and evaluate their students’ physical activity.
Sqord, Polar Active, Zamzee and others act as a bridge between the quantified self trend and physical education pedagogy, and represent a convergence of devices, software, apps, techniques and discourses of self-quantification with pedagogic practices, commercial imperatives and governmental health agendas. This hybrid mix of pedagogic technologies and modes of self-management is ordered and organized (at least partly) by underlying algorithms and their inbuilt models of the body in order to make the health of the child amenable to measurement and management, and constitute an emerging form of digitally mediated biopedagogy.
Importantly, devices such as these represent the increasing hybridity of insights about the body emerging from the biosciences with biosensing technologies that can detect signals from the body and transform it into measurable data, along with new biopedagogic techniques of bodily optimization. There is a sense in which the body of the child is understood as a kind of biological ‘software’ that can be debugged, patched and re-programmed—a kind of body-in-beta. The important object of inquiry into the biopedagogies of health-tracking and personal analytics devices then, should be in the network of relations between bioscientific explanations, algorithmic model-building and the design of the biopedagogic devices that produce the child as a body-beta.
Affective algorithms: Psychopedagogies of emotional maximization
‘Psychopedagogy’ refers to attempts to intervene in the psychological make-up of the learner, and with maximizing the emotions. We can think of initiatives which draw on popular theories from psychology and the other ‘psy-sciences‘ to promote happiness in education, meta-cognition, multiple intelligences, learning styles, emotional intelligence, and thinking skills, amongst others, as various forms of psycho-pedagogy. Many such ideas derived from the psy disciplines have been deployed in the development of educational technologies and other theories of digitally mediated learning, as well as much more widely in the proliferating ‘happiness industry’ through such devices as mindfulness apps, happiness indices, and measurements of well-being.
Significantly, there has been much recent interest in the idea of ‘emotional learning analytics’. Emotional learning analytics technologies are based on the collection and analysis, in real time, of data collected about children’s affective experiences in schools and online learning programmes. These technologies make extensive use of psychometrics, sentiment analysis and other measurable indicators of the emotions, in order to to enable the automatic detection, assessment and analysis of emotions. These can provide teachers with the necessary emotional information required to decide on appropriate pedagogic intervention, enable educational researchers to gain new insights into the affective dimensions of schooling, and enable school administrators and teachers to access dynamic real-time visualisations of pupils’ affective experiences. Emotional learning analytics involves techniques of content analysis, natural language processing, and the use of behavioural indicators, as well as quantitative instruments, qualitative approaches, well-being clouds, and intelligent tutoring systems.
Several biosensor and biometric technologies for measuring and monitoring the affective dimensions of the learner experience have been developed for schools, including ‘student sensor bracelets’ designed to detect excitement, stress, fear, engagement, boredom and relaxation through students’ skin, and facial recognition technologies powered by computer vision algorithms to produce automated metrics of student engagement. Technologies such as student sensor bracelets, which have been funded by the Gates Foundation, send a small current across the skin and measure changes in electrical charges as the sympathetic nervous system responds to stimuli. These galvanic skin response bracelets measure how well the skin conducts electricity, which varies with its moisture level; sweat glands are controlled by the nervous system so skin conductance can be used as an indication of emotional response. Particular models and understandings of the emotions and their measurability through biosensing devices underpin these applications. Psychological insights into eye movement, twinned with technical expertise in eye tracking, as well as the psychology of facial expression twinned with computer vision algorithms, represent the enmeshing of both ‘psy’ and ‘CompSci’ expertise in the enactment of the devices.
Significantly, emotional learning analytics and related applications depend on expert techniques that can classify human emotion. A recent review of emotional learning analytics research found reference to over 100 different measurable emotions, all linked to different techniques and methods of identification, classification and measurement, and concluded that ‘with increased affordances to continuously measure facial and voice expressions with tablets and smartphones, it might become feasible to monitor learners’ emotions on a real-time basis’. To understand new psychopedagogic applications such as student sensor bracelets and emotional learning analytics, it will be necessary to interrogate the underlying psy-sciences that inform their design, and to examine the ways in which the emotions can be classified and modelled in order to make them identifiable and measurable by an algorithm. The affective algorithms of emotional learning analytics work on new expert models and classifications of human emotion, and constitute a new mode of measurement and intervention in the psyche of the child.
Computing brains: Neuropedagogies of brain empowerment
An emerging development in educational technology is in ‘cognitive-based learning systems’ that are informed by neuroscientific methodological innovations, technical developments in brain-based computing, and neural networks algorithms. This is related to the growing discourse of ‘neuroscience in education,’ ‘neuroeducation,’ and ‘neuropedagogies.’ These terms articulate the mobilization of neuroscientific understandings of the learning process for the design and application of better pedagogies.
With the development of ‘cognitive computing’ through the R&D departments of organizations such as IBM, however, we are seeing the design of learning technologies running on neural networks algorithms and other brain-like forms of computation that are themselves modelled on particular brain-based cognitive processes. IBM has become a major organizational player in relation to cognitive computing in education. In particular, IBM is seeking to apply its own ‘brain-inspired’ advances in cognitive computing to the classroom. IBM has developed a specific ‘Cognitive Computing for Education Transformation’ strand of activities. Cognitive computing at IBM is a category of technologies intended ‘to enable people and machines to interact more naturally to extend and magnify human expertise and cognition.
IBM has been responsible for significant developments in neurosynaptic systems and neuromorphic hardware, sometimes referred to in promotional IBM literature as ‘computing brains,’ ‘systems that can perceive, think and act,’ or even a ‘brain-in-a-box’. Significantly, it also claims to be combining its neurosynaptic developments with neuroplasticity. Plasticity is the understanding that the brain’s neural architecture is itself pliable, flexible, and constantly adapting to environmental input; the brain is fundamentally a ‘learning brain’ constantly adapting to external stimuli. IBM’s engineers are attempting to model the neural plasticity of the ‘learning brain’ in silicon.
IBM’s cognitively ‘smarter classroom’ initiative is just one application of cognitive computing. In its imaginary of the classroom in five years, IBM grandly claims that the IBM ‘smarter classroom’ is a ‘classroom that will learn you’ through cognitive-based learning systems. The cognitive classroom promises personalization of the learning experience, real-time feedback on learner performance, adaptive learning software that can learn from and adapt to the learner, and intelligent software tutors that can automate remedial intervention or even prescribe appropriate curricular content.
IBM’s Cognitive Computing for Education Transformation program director has described such cognitive computing systems as intelligent, interactive systems:
where the computers have attributes that allow them to learn and interact with humans in more natural ways. At the same time, advances in neuroscience, driven in part by progress in using supercomputers to model aspects of the brain … promise to bring us closer to a deeper understanding of some cognitive processes such as learning. At the intersection of cognitive neuroscience and cognitive computing lies an extraordinary opportunity … to refine cognitive theories of learning as well as derive new principles that should guide how learning content should be structured when using cognitive computing based technologies.
Automated ‘cognitive learning content’, ‘cognitive tutors’ and ‘cognitive assistants for learning’ integrated into ‘personalized adaptive learning systems’, all ‘designed with a deep understanding of underlying cognitive neuroscience as well as cognitive theories of learning’, all exemplify how cognitive computing is imagined by IBM engineers as an intelligent computational augmentation to the learner’s cognitive process. One of these applications, the ‘cognitive tutor’, is even intended to:
supplement face-to-face teaching and ultimately replace it entirely for subjects and areas where a cognitive agent will, quite simply, do a better job of understanding the learner’s needs and provide constant, patient, endless support and tuition personalized for the user.
These developments instantiate the move of neuroeducation into the code/space of the digitized classroom, and prefigure emerging ‘neuropedagogies’ that need to be seen in terms of the data practices and algorithmic processing techniques as well as the neuroscientific brain modelling techniques that underpin them.
The development of cognitive learning systems, then, represent the enmeshing of ‘neuro knowledge‘ and its techniques of modelling the brain with the algorithmic expertise of cognitive computing development. Cognitive computing is entirely premised on the construction of ‘learning algorithms’ that, just like the plasticity of the ‘learning brain’ itself, are designed to learn by adapting to environmental stimuli and input. It is based on recent neuroscientific accounts of the ‘social life of the brain’ as a pliable organ characterized by its plasticity, and reflects the trend to apply neuroscientific insights as solutions to complex problems:
A range of new practices is emerging around the governing of human ‘embrained’ existence – new experts advising us how to live with, manage and improve our brains; … new modes of responsibilization urging individuals to care for their brain; and a new consumerization of the brain, offering us all manner of products, devices, exercises and the like to keep our brains healthy and maximize our brain power.
The neuroscientific insights emerging from IBM and being applied in its cognitive computing applications for the classroom need to be seen as socially situated products providing a particular working model of the brain that are accomplished through specific methodological measures and theoretical accounts of its internal functioning. Applications such as IBM’s cognitive tutor device need to be anatomized in terms of the neuroscientific models of the brain that underpin them, the ways neurological functioning have been modelled for algorithmic processing, and then the ways in which the applications themselves have been constructed to interact neuropedagogically with the learner to enhance their collective brain power as a symbiotic human-machine cognitive system.
Governing the data-child
In their book Governing the Child in the New Millennium published in 2001, Kenneth Hultqvist and Gunilla Dahlberg wrote about the ways in which children are the products of ‘historical truths’ that govern and guide their conduct. How children are conceived and practised upon are the products of particular ways of thinking which specify their assumed qualities and capacities.
Today, the emerging technologies and data practices associated with the biosciences of the child’s body, the psy-sciences of child emotions, and the neurosciences of child cognitive functioning and brain plasticity provide three dominant truths of the child that are being deployed to govern and guide them. The embodied, affective and embrained child is known through the datafied body, psyche and brain. These scientifically derived truths are finding their way into the design of technical systems and data practices that seek to measure the child in terms of bodily activity, emotional experience, and cognitive functioning, as bio-, psy-, and neuroscientific models on which algorithms can then be put to work. The task of governing the data-child is one in which the child, understood biologically, psychologically and neurologically, gets encoded through digital data practices that then ‘fold back‘ into new pedagogic techniques and devices that might actively alter the child’s bodily, emotional and cognitive conduct. Understanding these techniques of governing will require not just alarmist attention to the algorithms of data mining technologies. It will require careful anatomization of the entanglements of the sciences of the body, emotions and brain with algorithmic modelling practices and the application and enactment of new kinds of biopedagogies of body optimization, psychopedagogies of emotional maximization, and neuropedagogies of brain empowerment that together constitute the data-child.
code acts in education 