Computer Science · Security Unit · Session 7 of 8
Buying a security product is not a plan. Explaining your reasoning is.
Six sessions of individual controls now have to become one coherent document: a one-page security plan for the Harbour School Robotics Club, justified line by line against its actual risks — not a shopping list of "MFA, a firewall and encryption" with no explanation of why. This is where the unit stops being about naming controls and starts being about defending choices.

Executive summary
Goal: produce a practical plan, and justify every choice using the scenario itself.
| Task | Minutes |
|---|---|
| Recall controls and risks | 5 |
| Read the planning notes | 10 |
| Core video and question | 5 |
| Write your plan | 30 |
| Self-review and improvement | 10 |
1. Start with what must work
Write down assets and necessary activities before choosing technology. Members need their own payment status; they don't need everyone's contact details. Treasurers need to update payments; they don't automatically need to manage system accounts. The supervisor needs oversight and an agreed route to school IT.
2. Link every control to a reason
Reading the diagram: a plan is a reasoned cycle. Buying a security product is not the starting point, and writing the plan down is not the end of the process.
A useful sentence pattern:
Because [specific weakness] could allow [specific harm], use [control]. It works by [mechanism], protecting [security goal]. However, [limitation], so [complementary measure or justified trade-off].
3. Evaluate rather than list
For a small dataset, frequent full backups may be simpler to manage than a long incremental chain — a larger dataset might change that judgement entirely. More frequent backups reduce possible missing changes, but need suitable storage and management to match. Limiting permissions reduces opportunities for damage, but requires access to actually be updated when roles change.
Priority depends on context. Avoid inventing numerical likelihoods, and avoid claiming any one control solves every problem. Label your assumptions — for example, assume the school-approved records application supports individual permissions.
4. Make the plan testable
"Keep backups" is incomplete. Specify who checks success and who performs a restoration test. "Only treasurers can edit" should be tested with fictional member and treasurer roles in an authorised test environment, or described as a paper walkthrough — do not change actual school permissions for this lesson.
Use acceptance statements: "A member cannot edit another member's payment"; "The supervisor can recover a sample record within one school day"; "An outgoing treasurer's editing access is removed." These describe observable outcomes, not intentions.
Watch
Your main activity — One-page security plan
Self-review — 12-point learning checklist
Award 0 for missing, 1 for partial, 2 for clearly explained, in each category.
| Category | What earns 2 points |
|---|---|
| Risks | Three scenario-specific risks and a justified priority |
| Controls | Five complementary controls with mechanisms, including detection and recovery |
| Permissions | Roles have justified access; members do not receive unnecessary rights |
| Recovery | Frequency, protected location, responsibility and a test tied to the recovery target |
| Response | Prompt reporting and appropriate IT-led containment/investigation/recovery |
| Evaluation | Two meaningful trade-offs and two observable checks |
Aim for 10/12, then go back and improve your lowest-scoring category. This is a learning checklist, not an examination grade. A defensible plan might combine individual accounts with MFA, restricted editing, disk encryption, protected versioned backups, and monitored account/edit logs. It could specify nightly backups, a weekly success review and a monthly sample restoration, with frequency increased if a day's missing changes becomes unacceptable. These frequencies are example design choices, not universal rules. The teacher should own oversight and coordinate with school IT; students should not be assigned unsupervised system administration. A recovery-time target must be tested, not assumed from the backup schedule alone. A usable plan also names what happens when a treasurer leaves.Reveal model elements to compare after drafting
Key terms
- justification
- trade-off
- residual risk
- acceptance check
- separation of duties
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8