The command verb sets the standard
Every CIPD assessment criterion opens with a verb. It is not decoration — it is the thing being marked, and answering a level below it is the most common reason good work comes back referred.
CIPD assessment is criterion-referenced: the marker works down a list and decides, for each criterion, whether the evidence is there. Not whether the essay was good. Whether that specific criterion was met.
So a criterion that says evaluate, answered with a clear and well-written explanation, does not pass. The explanation might be the best paragraph in the submission. It still leaves the marker with a box they cannot tick, because evaluating means reaching a judgement and no judgement was reached.
That is usually what the feedback means when it says “more depth needed” or “insufficient evidence for AC2.3”. Not that you wrote too little. That you answered a different question.
explain
65 criteriaMake it clear how or why something works. Description is enough here — you are not being asked to judge it.
- 5CO01 AC2.1
- 5CO01 AC2.3
- 5CO02 AC1.3
- 5CO02 AC1.4
- 5CO02 AC3.2
- 5HR01 AC1.4
- 5HR01 AC3.1
- 5HR02 AC1.1
- 5HR02 AC1.2
- 5HR02 AC3.3
- 5HR02 AC4.2
- 5HR03 AC1.1
- 5HR03 AC2.1
- 5HR03 AC2.2
- 5HR03 AC2.3
- 5HR03 AC2.4
- 5HR03 AC3.3
- 5HR03 AC3.4
- 5OS01 AC2.1
- 5OS01 AC3.2
- 5OS01 AC3.3
- 5OS01 AC4.1
- 5OS01 AC4.2
- 5OS01 AC4.3
- 5OS02 AC2.3
- 5OS03 AC1.2
- 5OS03 AC2.2
- 5OS04 AC1.3
- 5OS04 AC2.2
- 5OS04 AC3.1
- 5OS04 AC4.2
- 5OS04 AC4.3
- 5OS05 AC1.2
- 5OS06 AC1.1
- 5OS06 AC1.2
- 5OS06 AC1.3
- 5OS07 AC1.1
- 5OS07 AC3.3
- 5OS07 AC3.4
- 3CO01 AC2.2
- 3CO01 AC3.1
- 3CO02 AC1.1
- 3CO02 AC1.2
- 3CO02 AC1.3
- 3CO02 AC1.6
- 3CO02 AC2.1
- 3CO03 AC1.1
- 3CO04 AC1.1
- 3CO04 AC1.2
- 3CO04 AC1.3
- 3CO04 AC1.4
- 3CO04 AC2.1
- 3CO04 AC2.4
- 3CO04 AC3.1
- 3CO04 AC3.2
- 3CO04 AC3.4
- 3CO04 AC3.5
- 3CO04 AC4.1
- 3CO04 AC4.2
- 3CO04 AC4.3
- 3CO04 AC5.1
- 3CO04 AC5.2
- 3CO04 AC5.3
- 3CO04 AC6.1
- 3CO04 AC6.4
assess
19 criteriaJudge the significance or extent of something. Close to evaluate, but the emphasis is on how much it matters.
discuss
18 criteriaSet out more than one view and show how they relate. A one-sided answer is the usual miss.
evaluate
17 criteriaWeigh it up and reach a judgement. Both sides, then a position — an answer with no verdict does not meet this.
analyse
6 criteriaBreak it into parts and show how they connect. Listing the parts is not analysing them.
summarise
5 criteriaThe main points, briefly, in your own words. Leaving things out is the skill being tested.
recommend
4 criteriaSay what should be done, and why that rather than the alternatives.
identify
4 criteriaName the specific ones that apply here. The lowest bar of any CIPD verb — but it still wants the actual items, not a description of the category.
consider
3 criteriaWeigh it up and say what you make of it. Lighter than evaluate, but still a view rather than a description.
compare
3 criteriaSet two or more things side by side. Similarities and differences, not one then the other.
examine
3 criteriaLook at it closely and in detail, drawing out what is not obvious at a glance.
conduct
3 criteriaCarry out the calculation or activity and show the working. The result on its own is not the evidence.
present
2 criteriaProduce it in a form a stakeholder could use. Format is part of the criterion.
appraise
2 criteriaJudge the value or quality of something, with reasons. Your view has to be visible.
recognise
2 criteriaShow you can identify it in a real situation, not just define it.
reflect
2 criteriaWrite about your own practice and what changed because of it. First person is correct here, and the outcome matters more than the account.
explore
2 criteriaLook at it from more than one angle and follow where it leads. Breadth counts for more than a verdict here.
distinguish
2 criteriaDraw the line between them clearly enough that a reader could sort a new example into the right one.
demonstrate
2 criteriaShow it happening through evidence of your own practice. This one cannot be answered in the abstract.
develop
2 criteriaProduce the thing. A criterion that says develop wants the artefact submitted, not a description of how one is made.
interpret
1 criterionSay what the data means, not what it says. Trends, causes, what follows.
make
1 criterionProduce the thing itself. “Make justified recommendations” wants the recommendations on the page, each with its reason — not an account of what a good recommendation would contain.
formulate
1 criterionProduce the actual thing — a plan, a set of activities — not a description of what one would contain.
differentiate
1 criterionShow what makes two things different, and why the difference matters. Defining both and stopping is the usual miss.
advise
1 criterionWrite as if to someone who has to act. A recommendation and its reasons, not a description of the topic.
design
1 criterionProduce the design itself, with the reasoning visible. A criterion that says design is not satisfied by explaining what a good one would look like.
review
1 criterionSurvey what is out there and say what you make of it. A list with no assessment is the usual miss.
define
1 criterionSay what it is, precisely enough to tell it apart from things it is often confused with.
specify
1 criterionName the particular things, not the general category. “I follow GDPR” is the category; which parts, and how, is the criterion.
participate
1 criterionYou have to have taken part, and there has to be evidence of it. Reading about the activity does not meet this.
describe
1 criterionSet out what something is like, in enough detail that a reader could recognise it. No judgement needed.
Check the verb before you check the word count
When a criterion comes back referred, the first thing to do is not to write more. It is to read the verb again, then read what you wrote, and ask whether the two match. Most of the time the fix is a paragraph that takes a position, not another five hundred words of context.
If you are staring at a referral and cannot see the gap, that is exactly the kind of thing a second pair of eyes settles in twenty minutes.
Stuck on something? Start a chat.
Tell us what you're working on and what's going wrong. We'll tell you what would help, what it costs, and how long it takes — before you commit to anything.
Replies in under 2 hours. No account needed to ask.