Database normalization for HKDSE ICT: 1NF, 2NF, 3NF without the confusion
What this topic covers
Normalization is the part of the Databases elective that most reliably separates students who’ve memorised a definition from students who actually understand what problem it solves. At its core, normalization is just a structured way of asking: is this table storing the same fact in more than one place, or storing more than one kind of fact in the same place? Everything about 1NF, 2NF, and 3NF follows from that one question.
Main areas
Why normalize at all. Before naming any normal form, it’s worth being able to explain the actual problems normalization prevents: update anomalies (changing a fact in one row but not another that stores the same fact), insertion anomalies (being unable to add one piece of data without also having an unrelated piece), and deletion anomalies (losing information you still needed, as a side effect of deleting something else).
First Normal Form (1NF). Every column holds a single, atomic value—no repeating groups, no multiple values crammed into one field. A table listing a student’s three phone numbers in one comma-separated column is the classic violation.
Second Normal Form (2NF). Builds on 1NF, and only becomes a meaningful question when a table has a composite primary key (two or more columns together forming the key). It requires every non-key column to depend on the whole key, not just part of it—this is where “partial dependency” comes from.
Third Normal Form (3NF). Builds on 2NF by removing transitive dependencies—a non-key column that depends on another non-key column, rather than depending on the key directly. If a table stores a student’s class and, separately, the class’s homeroom teacher, that teacher technically depends on the class, not on the student—a transitive dependency worth splitting out.
How the ideas fit together
Normalization connects directly to database design and to writing correct SQL later: a properly normalized structure makes most queries and updates straightforward, while an unnormalized one tends to produce exactly the anomalies normalization exists to prevent. It’s also one of the clearest places in the whole ICT syllabus where “understanding the concept” and “correctly applying it to an unfamiliar table” are genuinely different skills.
Common difficulties
The most common mistake is treating 1NF/2NF/3NF as a memorised checklist applied mechanically to a table, rather than as three specific dependency problems being tested for in sequence. Students often can recite “2NF means no partial dependency” but freeze when asked to identify a partial dependency in a table they haven’t seen before, because they memorised the definition rather than the reasoning behind it. Confusing partial dependency (2NF’s concern) with transitive dependency (3NF’s concern) is the single most common mix-up.
What you should be able to do
- Explain, in your own words, why an unnormalized table is a problem—citing a specific anomaly, not just “it’s bad practice.”
- Identify whether a given table is in 1NF, and correct it if not.
- Given a table with a composite key, identify a partial dependency and normalize it to 2NF.
- Given a 2NF table, identify a transitive dependency and normalize it to 3NF.
Related topics
- Web Application Development—where database design decisions show up again in a different context.
- Computational Thinking and Programming—the logical-reasoning skill normalization also draws on.
Get in touch
HKDSE ICT Tutoring