2–4 · Data, algorithms and digital power

IB Diploma Programme Digital Society HL (first assessed in 2024). The course looks at how digital systems 数字系统 change people's lives. It joins three parts: seven concepts 概念 (the questions you ask), seven content 内容 areas (the technology you study) and seven contexts 情境 (the parts of life where it matters). This topic covers the concepts and the first two content areas, data and algorithms. They are examined in Paper 1 (2 hours 15 minutes, 35% at HL; 1 hour 30 minutes, 40% at SL), where questions ask about the course and about real-world examples 真实例证.

In 2020 the COVID-19 pandemic cancelled England's A-level exams. Instead, a formula decided students' grades. It took each school's results from earlier years and each teacher's ranking of their students, and fitted this year's class to the old pattern. On results day, almost 40% of the grades that teachers had predicted were lowered. A strong student in a school with weak past results could not get the top grade, because the school "never" got it. Students protested in the streets with signs against "the algorithm", and four days later the government dropped the formula and used the teachers' grades instead. A few lines of rules had decided the futures of thousands of young people. This is what digital society studies.

Must-know for data, algorithms and digital power:

You do not study these 21 items one by one. Every inquiry takes one or two from each column and joins them to a real case.

A concept is a question you ask about a case:

Data 数据 are raw facts: numbers, words, times, places. Information 信息 is data that has been organised so that it means something. A smartwatch records your heart rate as data (72, 75, 140, 138...); "your heart rate rose sharply during your run" is information.

Every dataset has makers, who decided three things:

• Collection 收集: what is recorded, and what is left out. A ranking records the teacher's order of students, not how hard each one worked during the pandemic.
• Processing 处理: cleaning, joining and analysing the data.
• Storage 存储: who keeps the data, where and for how long.

Data about people raises questions of privacy 隐私. In 2014 a personality-quiz app on Facebook was used by about 270,000 people, but it also collected data about their friends. The data of up to 87 million people reached the political consulting firm Cambridge Analytica, which used it to target political advertising. Most of those people had never used the app. In 2019 the US Federal Trade Commission fined Facebook US$5 billion for its privacy failures. Data collected for one purpose (a fun quiz) was used for another (political campaigns).

An algorithm 算法 is a set of step-by-step rules that turns inputs into outputs, like a recipe. Algorithms decide what appears in your video feed, which route a map app suggests and whose job application is read.

The key idea: bias in, bias out. An algorithm learns from or uses data from the past, so it can repeat the unfairness of the past. Two examples:

• The 2020 grading formula used each school's past results as an input, so students at schools with weaker histories were held down, whatever their own work.
• Amazon tested a computer system for choosing job applicants, trained on ten years of CVs, most of them from men. It learned to give lower scores to CVs that included the word "women's" (as in "women's chess club"). Amazon stopped using it (Reuters reported this in 2018).

A good inquiry joins all four parts: a concept, a content area, a context and a real-world example. Together they make a focused question.

Worked example. Power (concept) + algorithms (content) + social and political contexts + the 2020 UK grading formula = "Did the formula move the power to decide a grade from teachers to the formula's designers?"

Three common mistakes:

• Concept only: an essay about power in general, with no real case.
• Case only: the story of 2020, with no concept to explain it.
• Context dump: a lot about schools and politics, but no digital system in view.

Paper 1 questions range from short answers to extended responses. The IB's own examples:

For a long answer: define the key terms; name the concept, content and context you are using; explain one or two real-world examples in detail (who, where, when, what happened); give more than one perspective; and end with a judgement.

Worked plan. Describe a context and real-world example in which collecting data about online activity may pose privacy concerns for young people.

• Context: political (or social).
• Example: the Cambridge Analytica case: a quiz app collected data about its users' friends, who never agreed.
• Concern: young people share likes and friend lists freely; they may not know that the data can be moved and used to target them.
• Link: data collected for one purpose was used for another (data, 3.1), which is a question of power and of values and ethics.

• Concepts, content and contexts: seven of each, joined through inquiry.
• Each concept as a question, applied to the 2020 grades.
• Data and information; collection, processing, storage; privacy and Cambridge Analytica.
• Algorithms as rules; bias in, bias out; the grading formula and Amazon's hiring system.
• The four-part focused question and the three common mistakes.
• Paper 1 command terms and the worked plan.