Time: about 14 minutes. Read page.md first.
Part 1 — same class, different amount (5 minutes)
Here are three tasks that all sit in the same class — each is a claim about the world, checkable only against a source.
| The task | |
|---|---|
| A | Confirming the opening hours of a library before you walk there |
| B | Confirming a submission deadline before you plan your week around it |
| C | Confirming a dosage interval that will be written into care notes for someone else |
For each, write three short answers:
- Kind of check — what would you compare it against? (This should be nearly the same answer for all three, and that is the point.)
- Amount — consequence, recoverability, and whether you could assess it.
- What you would actually do, in one line.
The exercise is to see the first column stay constant while the third changes completely. If your answer to question 1 varies a lot between the three, look again — you may be letting importance decide the kind, which is the error Unit 3 warned about.
Part 2 — the row you talk yourself out of (4 minutes)
Find a real case from your own study or work that belongs in the table's bottom row: large consequence, and you could not confidently assess a wrong answer.
Write four lines.
| The claim | |
| Why the consequence is large, and who bears it | |
| Why you could not assess it | |
| Who is qualified to settle it, and is that person reachable |
If you genuinely cannot find one, use this instead: think of something you did accept without being able to check it, and write the same four lines about that. Almost everyone has one, and it is more useful than an invented case.
Part 3 — write the handover (5 minutes)
Take the case from Part 2 and write what you would actually send. Four parts, as the reading gives them:
- The claim you need settled, stated precisely.
- What you have already checked, and what it showed.
- What you could not check, and why.
- What decision is waiting on it.
Then apply one test:
Have I handed over a question, or have I handed over an output and asked someone to deal with it?
The second wastes the other person's time and hides what you already know. The first is a colleague asking a colleague something.
Keep Part 3. It is the same artefact as the Unit 17 evidence pack, used for its other purpose: not to defend a claim you are making, but to hand over one you are not prepared to make.
Now take the checkpoints
cp12-two-axes is about Part 1.
cp12-escalation-trigger is about Part 2.
python3 selfcheck.py run cp12-two-axes
python3 selfcheck.py run cp12-escalation-trigger # or in the browser
Run the Python route from course/lab/, or use the browser self-check.
Optional extensions
Optional. These sit outside the core study time, are not required for the checkpoints, and nothing later in the module depends on them.
Find where the axes are already separated for you (about 5 minutes). Many professions have this built in — sign-off, second-checking, four-eyes rules, supervision requirements. Find one in a field you know and write down what it treats as the trigger. You will usually find it is consequence and irreversibility, not the likelihood of error, which is the same conclusion this unit reaches from a different direction.
Rewrite a refusal (about 4 minutes). Find a time you declined to answer something, or should have. Write it as a handover rather than a refusal: the claim, what you checked, what you could not, and what is waiting. The difference between "I do not know" and a handover is roughly four sentences, and it is most of the difference between being unhelpful and being useful.
Marking guidance — open this once you have done the activity, to check your own work
Read this after attempting the activity.
Part 1 — what the three tasks show
The kind of check is the same for all three. Each is a claim about the world with a source that exists: the library's own posted hours, the official assessment timetable, the prescribing reference. In Unit 1's terms all three are recall unless you supply the source; in Unit 3's terms all three need source checking. Nothing about consequence changes that.
The amount is completely different.
| Consequence | Recoverable | Could you assess it | What follows | |
|---|---|---|---|---|
| A | You walk somewhere for nothing | Yes, immediately | Yes | One look at the official page. Done. |
| B | You mis-plan a week; possibly a late submission | Partly, and expensively near the deadline | Yes | Check the official timetable, not a summary of it, and note the date you checked. |
| C | Harm to a person who is not you | No | Almost certainly not | Escalate. This is the bottom row. |
Three things worth noticing.
Row C is not distinguished by being "medical". It is distinguished by the three factors: the cost lands on someone else, it cannot be taken back, and you are outside your competence. A logistics figure with the same three properties gets the same answer.
Row B is the interesting one. It looks routine and is the one most people under-check, because the consequence is entirely to them and feels manageable. Note that its recoverability decreases with time, which is a property worth looking for generally.
"Could you assess it" is about you. Two people can get different correct answers for row C — a prescriber and a student are not in the same position — and the table is doing its job when that happens.
Part 2 — the row people talk themselves out of
The conversion from "large consequence, cannot assess" to "large consequence, can probably assess" happens in one of three ways, and they are worth recognising:
- Familiarity read as competence. You have seen a lot of writing on this, so the answer feels checkable. Recognising the vocabulary is not the same as being able to tell a right answer from a plausible one.
- Partial checks read as whole ones. You confirmed the parts you could, and the confirmed parts stand in for the rest. Unit 9's scoped statement exists to stop exactly this.
- The check that is not independent. You looked it up somewhere that is plausibly where the claim came from in the first place. Agreement between an output and its own likely source is not confirmation.
If you used the "something I did accept" variant, that is the better version of this exercise. A case you actually let through tells you which of the three above is your habit.
Part 3 — question or output
The test has a clear signal. A handover that begins "Can you check this for me?" and pastes an output is handing over your whole problem including the parts you had already solved. A handover that begins "I need to know whether X, and here is what I have established" is a question, and it can be answered in minutes.
The second also does something the first does not: it makes your own work visible, including its limits. That is why it is the same artefact as the evidence pack. The pack defends a claim you are making; the handover explains a claim you are declining to make. The structure is identical because both are answers to "what is established, and how?"
The checkpoints
cp12-two-axes tests the order of the two questions. The tempting wrong answers
let consequence decide the kind of check, which is precisely the error Unit 3
guarded against and which this unit is careful not to reintroduce.
cp12-escalation-trigger tests what actually licenses stopping. The trap is
subject matter: the triggers are about your situation — what you can check, what
you can undo, who bears the cost — not about the topic being a serious one.
Your private activity record
Browser storage is not a permanent copy
Progress is kept only in this browser, profile and device. Private browsing, clearing site data, removing the profile, a browser reset, storage eviction or device loss can erase it. Keep important answers and contributions separately.
These notes stay in this browser unless you download a backup or activity log.