DPHU1008 Chap.9 Datafication, Algorithms and Dataveillance
Datafication, Algorithms and Dataveillance
The final lectures follow data into society. Lecture 9 covers big data, datafication and data algorithms: how everyday activity becomes recorded traces, how value is extracted from them, and how they shape work and governance.
Lecture 10 connects data to surveillance and power, asking how machine learning, automation and prediction sit in everyday life, and Lecture 11 looks to data futures before a final consultation lecture for Assessment 3.
The required readings are boyd and Crawford's critical questions for big data, Viljoen on data relations and Jackie Wang on predictive policing.
Earlier readings already contain the patterns: classifiers that revive race science, watchlists triggered by a matching name, and platform ratings that discipline workers while appearing objective. Dataveillance, Roger Clarke's term, names monitoring people through their data trails.
For Assessment 3, the final lecture is a consultation that helps you fold these concepts into your reflection.
For the critical journal and the lens evaluation, datafication and dataveillance work well with everyday renderings: a fitness ranking, a loyalty app, a school attendance dashboard. Ask what each system records, what it predicts, what it cannot see and who carries the cost of its errors.
What this chapter covers
- 01
Datafication: everyday activity turned into recorded traces
- 02
Big data and the value of prediction across many people
- 03
What algorithms can and cannot know
- 04
Dataveillance: monitoring through data trails
- 05
Feedback loops in predictive policing
- 06
Critical questions for big data claims
- 07
Data futures: threats and community uses
- 08
Viljoen, boyd and Crawford, and Wang as the closing readings
- 09
Data relations: value and harm shared across people
Worked example · free
Questioning a workplace productivity score
- +1Datafication: every scan turns movement and time into a trace that becomes the measure of a worker.
- +1Algorithm: the score ranks traces against a target but cannot see injury, training or broken equipment.
- +1Dataveillance: workers are monitored continuously through their data trail, not by a supervisor watching.
- +1Power: flagged workers face discipline from a number that looks objective, echoing the platform ratings Jones describes.
Key terms
- Datafication
- The conversion of everyday activities into recorded data that can be stored and analysed.
- Big data
- Very large and varied pools of data traces combined for analysis and prediction.
- Algorithm
- A defined sequence of steps a computer follows to turn inputs into an output such as a ranking.
- Dataveillance
- Clarke's term for monitoring people through their data trails rather than direct observation.
- Feedback loop
- A cycle in which a system's outputs generate the data that confirm its next outputs.
Datafication, Algorithms and Dataveillance FAQ
Why can a predictive policing model be accurate and unfair at once?
Its training data record where police patrolled, not where all crime occurred. The model predicts future patrols well while sending officers back to the same neighbourhoods, so its accuracy reflects past policing rather than public safety.
What questions do boyd and Crawford raise about big data?
They argue that big data changes what counts as knowledge, that claims to objectivity are misleading, that bigger data are not always better, that data lose meaning out of context, and that accessible data are not automatically ethical to use.
Is the course only negative about data futures?
No. Earlier readings show communities counting themselves and Indigenous nations using mobile tools to collect their own data. The final lectures ask you to weigh risks and possibilities and fold both into your Assessment 3 reflection.
Is a feedback loop the same as bias in the data?
Not exactly. Biased data are skewed from the start; a feedback loop is a process in which a system's own outputs generate the data that confirm it. A loop can turn a small initial bias into a large and self-confirming one over time.
What does Viljoen mean by data relations?
Viljoen argues that data matter because they connect people to one another: what is learned from some people is used to make predictions about others. Harm and value can therefore be collective, so individual consent alone rarely settles the ethics of a data set.
Assessment move
Map your own data trail for one day: list each app or card that recorded you and what it might predict. Choose one system and draw its feedback loop in four boxes. Read the required essays alongside a current news story about an algorithm, and practise one paragraph that moves from what the system sees to what it misses to who bears the cost.
Read one recent news report about an algorithm making a decision and trace its inputs, its output and its possible feedback loop. After each lecture, note one system you use daily that fits that week's concepts.
DPHU1007 Media, Society and Politics · DPHU1001 Journalistic Voices and Values · DPHU1003 News Fundamentals
