VCE Vault

Year 9 Digital Technologies · Data and Information · 18m read

Data — Collecting, Storing and Analysing

Ever wonder why your phone seems to know what you're thinking about? Data collection is happening everywhere—through apps, websites, and even your location. You'll explore how organisations gather information, keep it safe in databases, and use analysis to spot patterns and make decisions. Learning this matters because you need to understand your privacy rights, protect your personal information, and make smart choices about what you share online.

Learning objective: Explain methods of data collection, storage, and analysis, including privacy considerations.

By the end of this lesson you can:

  • I can describe structured and unstructured data.
  • I can explain how databases store and retrieve data.
  • I can identify privacy risks in data collection.

Data is raw ore. Collected data is like rock from a mine. Cleaning, storing and analysing it refines it into valuable information.

The life of data

DATA TYPES:
• STRUCTURED: organised in rows/columns (databases, spreadsheets). Easy to search and analyse.
• UNSTRUCTURED: free-form (images, video, text documents, social media posts). Harder to analyse — requires AI/ML.
• QUANTITATIVE: numbers (age, temperature, price).
• QUALITATIVE: categories or descriptions (colour, opinion, name).
• METADATA: data about data — file size, creation date, location tags on photos.
• BIG DATA: datasets so large and complex that traditional tools can't handle them. Characterised by the "3 Vs": Volume, Velocity, Variety.

### Databases in Depth

RELATIONAL DATABASE:
• Data in tables; tables linked by "keys."
• SQL (Structured Query Language): used to query databases.

Common SQL statements:
SELECT name, year_level FROM students WHERE year_level = 9;
INSERT INTO students (name, year_level) VALUES ('Alice', 9);
UPDATE students SET year_level = 10 WHERE name = 'Alice';
DELETE FROM students WHERE name = 'Alice';

  • PRIMARY KEY: unique identifier for each record.
  • FOREIGN KEY: connects records in different tables.
  • NORMALISATION: organising data to reduce redundancy (1NF, 2NF, 3NF).

### Australian Data and Privacy Context

| Law / Standard | What it does |
|----------------|-------------|
| Privacy Act 1988 | Governs collection, use, storage of personal information in Australia |
| Australian Privacy Principles (APPs) | 13 binding principles for APP entities |
| Notifiable Data Breaches (NDB) scheme | Organisations must notify OAIC + individuals of eligible breaches |
| GDPR (EU) | Stricter EU regulation; applies to Australian businesses with EU customers |

DATA COLLECTION METHODS:
• Sensors, forms, surveys, web scraping, social media APIs, transactional records.
• BIAS in data collection: if data doesn't represent the full population, conclusions will be biased.
• PRIMARY DATA: collected firsthand for a specific purpose (your survey).
• SECONDARY DATA: collected by others; may be repurposed (ABS census data).

DATA LIFE CYCLE:
1. Collect → 2. Store → 3. Process/Analyse → 4. Share → 5. Archive/Delete

DATA MINIMISATION: only collect what you actually need — reduces breach risk.
COOKIES: small files tracking your behaviour across websites.
TRACKING PIXELS: invisible images in emails/web pages that record when opened.

AARI TIP: Ask Aari to explain what data Google or Facebook collects about you and why.

Data essentials

  • Primary key: Unique ID for each database record — no duplicates.
  • SQL SELECT: Queries data: SELECT column FROM table WHERE condition.
  • Privacy Act: Australian organisations must follow 13 Privacy Principles.

Worked example (Medium)

A school wants to store student data. Design a simple database structure and identify two privacy considerations.

  1. Tables. STUDENTS (StudentID, FirstName, LastName, YearLevel, DateOfBirth). SUBJECTS (SubjectID, SubjectName). ENROLMENTS (EnrolmentID, StudentID [FK], SubjectID [FK]).
  2. Keys. StudentID is the primary key for STUDENTS. EnrolmentID links the tables — allows one student to be enrolled in many subjects.
  3. Privacy 1. Sensitive data (medical, behavioural) should be in a separate, access-restricted table — need-to-know basis only.
  4. Privacy 2. Data must be stored securely (encrypted at rest) and only kept as long as needed — not retained indefinitely after students leave.

Answer: Good database design separates concerns, uses proper keys and builds in privacy controls from the start.

Common mistakes

  • Treating "data" and "information" as interchangeable Data is raw (42, "Melbourne", 2024-04-01); information is data in context with meaning ("42 students enrolled in Melbourne in April 2024") Fix: Define each clearly — data becomes information when it is organised and given context
  • Thinking a foreign key must be unique Primary keys must be unique; foreign keys can repeat (many students enrolled in the same subject) Fix: A foreign key references a primary key in another table — it can appear multiple times in the linking table
  • Saying organisations can keep your data forever APP 11 requires organisations to destroy or de-identify data when it is no longer needed for its purpose Fix: Data retention must be limited and purposeful under the Privacy Act 1988

Worked example (Hard)

Design a normalised relational database for a school library system. Describe the tables, keys and at least two SQL queries to retrieve useful information.

  1. Identify entities. A library system involves: Books, Members (students/staff), and Loans (borrowing records). Each is a candidate for its own table.
  2. Table design. BOOKS (BookID [PK], Title, Author, ISBN, Genre, CopiesTotal) MEMBERS (MemberID [PK], FullName, YearLevel, Email) LOANS (LoanID [PK], BookID [FK], MemberID [FK], DateBorrowed, DateDue, DateReturned)
  3. Normalisation check. Each table holds data about one entity — no repeated groups. If a book has multiple authors, we would need an AUTHORS table and a BOOK_AUTHORS join table to avoid storing multiple authors in one column (violates 1NF).
  4. SQL Query 1. Find all books currently on loan (not yet returned): SELECT b.Title, m.FullName, l.DateDue FROM LOANS l JOIN BOOKS b ON l.BookID = b.BookID JOIN MEMBERS m ON l.MemberID = m.MemberID WHERE l.DateReturned IS NULL;
  5. SQL Query 2. Find members with overdue books: SELECT m.FullName, m.Email, b.Title, l.DateDue FROM LOANS l JOIN MEMBERS m ON l.MemberID = m.MemberID JOIN BOOKS b ON l.BookID = b.BookID WHERE l.DateReturned IS NULL AND l.DateDue < CURRENT_DATE;
  6. Privacy consideration. Member email and borrowing history is personal information under the Privacy Act. Access should be restricted to library staff only. Loan history should be deleted when a member leaves the school (APP 11 — data retention).

Answer: A well-normalised database avoids redundancy, uses appropriate keys, and builds in privacy controls. SQL JOINs allow powerful cross-table queries.

Practice questions

  1. In a relational database, a "primary key" is:

    1. The most important piece of information.
    2. A unique identifier for each record in a table.
    3. The first column in a table.
    4. A password protecting the database.

    (1 marks)

  2. Explain how bias in data collection can lead to unfair outcomes in AI systems.

    (3 marks)

  3. The "Notifiable Data Breaches" scheme requires Australian organisations to:

    1. Report all data breaches to the police.
    2. Notify the OAIC and affected individuals when a breach is likely to cause serious harm.
    3. Delete all customer data after a breach.
    4. Pay a fine to affected customers.

    (1 marks)

  4. A health app company collects user location, health metrics, and browsing history. Evaluate the privacy risks and describe four actions the company should take to comply with Australian Privacy Principles.

    (4 marks)

  5. Explain the difference between structured and unstructured data with one example of each.

    (2 marks)

Deeper Understanding: Big Data and the Data Economy

The volume of data created globally has grown exponentially — in 2023, approximately 120 zettabytes (120 trillion gigabytes) of data existed worldwide. This "Big Data" era is characterised by the 5 Vs: Volume (sheer amount), Velocity (speed of generation and processing), Variety (structured, unstructured, semi-structured), Veracity (reliability and accuracy), and Value (the business/social insight extractable).

Australia's data economy is substantial. The Australian Bureau of Statistics (ABS) collects and publishes data on virtually every aspect of Australian life — the Census, CPI, unemployment, health statistics. This data underpins government policy, business decisions, and academic research. Under Open Data policies, much of it is freely accessible at data.gov.au.

The tension between data utility and privacy is one of the central technology policy debates of our time. "Data linkage" — combining datasets from different sources to create richer pictures — is enormously powerful for public health research (identifying disease patterns across hospital, Medicare and Census data) but creates profound privacy risks. Australia's My Health Record system and the 2016 Census de-anonymisation controversy illustrate how difficult true anonymisation is when multiple datasets are combined.

The concept of data sovereignty is increasingly important, especially for Aboriginal and Torres Strait Islander communities who assert the right to control data about their communities, land and culture. The AIATSIS Code of Ethics for Aboriginal and Torres Strait Islander Research establishes principles for ethical data collection and ownership in Indigenous research contexts — a model being studied internationally.

Summary

  • Data types: structured vs unstructured; quantitative vs qualitative; primary vs secondary.
  • Relational databases use tables, primary keys and foreign keys; SQL queries retrieve and manipulate data.
  • The Privacy Act 1988 and 13 Australian Privacy Principles govern data handling in Australia.
  • The Notifiable Data Breaches scheme requires organisations to report serious breaches to OAIC and individuals.
  • Data minimisation — collecting only what you need — reduces privacy risk.
  • Bias in training data creates biased AI outcomes, requiring diverse data collection and algorithmic auditing.

More from Year 9 Digital Technologies