How to Write Tests of Control in ACCA AA

I want to start with the most valuable sentence you'll read in your AA preparation: "A test which starts with 'check' is unlikely to provide sufficient detail." That's a direct quote from the ACCA AA examiner — and it appears, almost word for word, in the MJ24, SD24, and D25 examiner reports. Three consecutive sessions, same feedback. Which means thousands of students across India, Asia and globally are still writing "check that purchase orders are signed" and wondering why they lose the mark. This guide is going to fix that permanently. I'll show you exactly what the examiner wants, give you the five verbs that consistently score, and walk you through before-and-after examples from all three main business cycles. By the time you finish reading, you'll have a template you can apply to any test of control question, any sitting, any scenario. If you want the broader picture first, our audit risk step-by-step guide explains how audit risk identification flows into tests of control — the two topics are tested together in most Section B questions.

Why "Check" Doesn't Work — The Examiner's Exact Words

Let me be precise about what the examiner is saying here, because there's a specific reason "check" fails — and once you understand it, you'll never use it again.

The word "check" is a summary, not a procedure. When you write "check that purchase orders are authorised," you've told the examiner what outcome you're looking for — but you haven't told them how the auditor will actually do it. Will the auditor inspect a document? Ask someone? Re-perform the calculation? Watch the process? Each of those is a different procedure with a different level of evidence reliability. The examiner can't award the mark because they don't know what you're doing.

Here's the second thing "check" fails to do: it doesn't describe the document, the sample, or what you're looking for as evidence that the control has operated. The MJ24 examiner made this point with purchase order authorisation: the correct answer wasn't just "inspect purchase orders for signatures" — it was "inspect a sample of purchase orders for evidence of authorisation by the relevant individual in line with the established authorisation levels." That final clause — the control objective — is what converts a ½-mark answer into a full-mark answer.

The Anatomy of a 1-Mark Test of Control (4-Part Formula)

Every full-mark test of control answer has four components. Miss one and you're at ½ mark. Miss two and you may get nothing. This is the same specificity principle that applies throughout AA — our internal control deficiency guide shows how the same precision is required for every part of the control deficiency question, including the test of control at the end.

Let's apply this to a simple payroll control — say, "payroll is authorised by the HR director before payment is processed."

The difference between ½ mark and 1 mark is the control objective clause at the end. You need to say what you're looking for as evidence — not just that you inspected a document. Why must it be the HR director specifically? Because that's the control. Why must it be prior to payment? Because that's when the control is supposed to operate. The examiner rewards you for demonstrating that you understand what the control does, not just what document to look at.

5 Verbs That Score Marks Every Time

Your verb choice does two things: it defines the type of evidence you're obtaining, and it signals to the examiner that you understand audit procedure methodology. Here are the five that consistently score, pulled directly from CIA Part 2 Section 2165 and confirmed by the AA examiner mark schemes.

What NOT to Do: Observe, Substantive Procedures & Management Actions

Knowing what to avoid is as important as knowing what works. These three failure modes cost students significant marks — and two of them appear explicitly in examiner reports.

❌ Failure Mode 1: Over-relying on "Observe"

The MJ24 examiner stated clearly: "Tests such as 'observe' should be used sparingly as they do not generally score as well as inspection or some types of enquiry procedures." Observation is only appropriate when there is no documentary alternative — for example, observing the physical inventory count or observing the supervision of employee ID card sign-in. In almost every other case, a document exists that you can inspect. Use inspection. Observation as a default answer suggests you're not thinking about what evidence is available — and the examiner notices.

❌ Failure Mode 2: Writing a Substantive Procedure Instead of a Test of Control

This is the most penalised mistake across all AA sessions. Our guide on the exact difference between substantive procedures and tests of control explains this in full — including real examples from the D25 exam session. The key principle: a test of control provides evidence that the client's control has operated. A substantive procedure provides evidence about whether a financial statement balance is correct. They can use the same document, but the purpose is completely different.

❌ Failure Mode 3: Writing a Management Action Instead of an Auditor Procedure

Tests of control are auditor procedures — things the auditor does to gather evidence. The D25 examiner specifically called out tests that read as management recommendations rather than auditor actions. "Ensure that segregation of duties is in place" is a management instruction. "Inspect a sample of bank reconciliations and confirm the preparer and reviewer are different individuals by checking signatures against the HR team list" is an auditor procedure. Your test must start with what the auditor will do — not what management should do.

The CIA Part 2 Foundation That Gives You the Edge

Students who come to ACCA AA having studied CIA Part 2 have a structural advantage on test of control questions. Here's why — and how to get the same advantage even if you're only sitting AA.

CIA Part 2 Section 2162 also covers the distinction between compliance testing (what AA calls tests of control) and substantive testing — the exact distinction that costs so many AA students marks. Once you've studied it from the CIA perspective — which frames it as "process-oriented vs. data-oriented" — you stop confusing the two permanently. Our internal audit process guide covers the fieldwork phase in detail, explaining how evidence gathering — inspection, observation, inquiry, reperformance — directly maps to what AA tests of control questions are asking for.

The CIA is recognised in 170+ countries — including strong and growing communities across India and the Gulf — and increasingly valued by employers precisely because it demonstrates this kind of methodological rigour. For candidates studying for both qualifications, the CIA Course by Surgent on eduyush covers all three parts with adaptive question technology. And if you want to understand how CIA and AA connect at the exam-syllabus level, the CIA exam structure guide maps it clearly.

Before & After: Sales Cycle Controls (3 Examples)

The sales cycle is the most frequently examined cycle in AA. Controls here typically involve authorisation of orders, matching of documents, and review of credit limits. Here are three controls with the full progression from poor to full-mark.

Before & After: Purchase Cycle Controls (3 Examples)

Purchase cycle controls typically involve purchase order authorisation, three-way matching, and supplier approval. The MJ24 examiner's worked example — purchase order authorisation by relevant officials — is a classic, and I'll use it here exactly as the examiner described it.

Notice the pattern across all three: every full-mark answer has a specific verb, a named document, "a sample of" to signal operating effectiveness testing, and a clause explaining what evidence would confirm the control is working. That last clause is where most ½-mark answers become 1-mark answers. Practicing with BPP ECR past papers with model answers lets you compare your wording directly against full-mark examples and train this instinct.

Before & After: Payroll Cycle Controls (2 Examples)

Payroll controls are tested almost every session. The D25 examiner report specifically used a payroll example to explain the failure mode — so these examples are especially worth studying carefully. Authorisation of payroll, segregation of duties, and new starter/leaver controls are the most common payroll controls examined.

The authorisation and fraud prevention connection is important here: missing authorisation controls create the "opportunity" component of the fraud triangle. Understanding why the control exists — to prevent ghost employees or fictitious payroll entries — helps you write the control objective clause correctly, because you can explain what the evidence confirms.

The Copyable Template — Use This in Every Exam

Take this template into every AA sitting. It works for every cycle, every control type, every session.

Applied: "Inspect a sample of purchase orders and review for evidence of authorisation by the relevant official in accordance with established authority levels, to confirm that all purchases have been appropriately approved before being placed."

This template also integrates naturally with the control deficiency structure in Section B. Once you've identified the deficiency, recommended a control, and described the test, your complete answer for one deficiency follows: identify → explain implication → recommend control → describe test. All four parts use the same principle of naming specifics. See our complete control deficiency guide for the worked examples showing all four parts together. And once you're confident on tests of control, the natural next question is how they differ from substantive procedures — our substantive procedures guide has the clearest explanation you'll find, with 2025 exam examples.

What We See on Exams — Faculty Observations

Common Mistakes We See — And How to Fix Them

Frequently Asked Questions

Recommended Resources for Test of Control Mastery