
Direct answer: Use these AUD sampling flashcards as evidence drills. For every item, write four boxes: the population, the selection or analytic signal, the exception or limitation, and the person who still owns the conclusion. Then preserve the next record that could confirm your response. The card is useful only when it changes what you review next.
Scope and freshness note: This is original U.S. CPA study material, not an exam-authority publication. Reopen the AICPA CPA Exam Blueprints for the active testing window, the AICPA AUD resources for current subject boundaries, and the NASBA FAQ for jurisdiction, eligibility, scheduling, or score questions.
Use the AUD review hub to orient the section. Then choose the Blueprint format guide for task boundaries, the topic MCQ page for quick retrieval, or the TBS practice guide when an exhibit or procedure is the missing skill.
For a broader preparation path, continue with our related CPA study resources.
AUD sampling flashcards: build an evidence card before the answer
Sampling, data analytics, and specialist work can look different on the screen while testing the same AUD habit: match the evidence to the claim. First, a population defines what could be represented. Next, a selection shows what was examined. An analytic result prioritizes records but may not establish an exception. A specialist report adds technical work with a stated scope. A correct response keeps those roles separate.
To begin, write the task verb beside the four boxes. “Identify” asks you to locate the controlling fact. “Evaluate” asks you to judge evidence quality or limits. “Determine” asks for a response supported by the record. “Recommend” asks for the next action. The word alone never proves the conclusion; it tells you what your note must explain.
Use a 9-minute AUD sampling card
For timing, give a new item two minutes to read, three minutes to choose and defend an action, and four minutes to review the boundary. When the timer ends, stop adding facts. Put a star beside the first box you could not complete. In addition, a short, honest gap is more useful than a long paragraph that hides the missing record.
AUD sampling evidence card: four boxes
Box 1 — Population: name the records that belong, the period, and the cutoff. “Claims exported April 2” is a data event; it is not automatically the complete March population.
Box 2 — Selection or signal: state how a row was chosen or why an analytic rule surfaced it. A random sample, a duplicate-number rule, and a specialist model answer different questions.
Box 3 — Exception or limitation: write the fact that can change the response. Late postings, missing identifiers, an untested assertion, or an assumption outside scope belongs here.
Box 4 — Responsibility and record: identify who must evaluate or obtain the remaining evidence, then name one document you would save. This prevents a technical report or a dashboard total from silently becoming someone else’s conclusion.
In practice, keep one sentence per box. If a box needs a paragraph, circle the first sentence that still depends on an unverified assumption. That sentence becomes the target of the next review action.

Run the original AUD sampling flashcards
Before opening a rationale, attempt the scenarios. Write the four boxes and choose the answer that protects the evidence boundary. These are original practice cases for a U.S. CPA study session; they are not official, recalled, leaked, or exam-memory questions.
Reveal the AUD reasoning
Attempt the original item first. Write the controlling fact, the boundary or evidence check, and the reason you would save before opening the explanation.
Reveal the answer
Answer: Compare your note with the rationale, mark the first missing proof point, and change one fact before retrying. Do not use a raw answer letter as the study record.
Compare your process with the reasoning shown above.
Case 1: West Harbor Medical and the late claims frame
Fact pattern: For example, West Harbor Medical freezes an outpatient claims export at 8:00 a.m. on April 2. The dashboard total agrees to the file. At 1:15 p.m., two reversal batches post for March services. The proposed sample was drawn from the morning export.
Question: What should the learner resolve before using the sample to support the March population?
- Accept the sample because the dashboard and export agree.
- Call the reversals immaterial because the selected claims look ordinary.
- Rebuild the period population through the stated cutoff, include the late batches, and reassess whether the selection represents the claim.
- Abandon sampling whenever a system has a daily freeze.
Answer and rationale: Choice 3 is strongest. Agreement between the views proves that they share a morning total; it does not prove that the March frame includes records posted later. Option two treats an ordinary subset as evidence about excluded rows. Option four replaces a limitation with an absolute rule. Record: the export timestamp, reversal register, and revised frame.
Case 2: Mesa Ridge Components and the analytic cue
Fact pattern: Similarly, Mesa Ridge Components runs a rule that flags invoices with the same vendor, invoice number, and amount. Fourteen pairs appear. The accounting extract contains no shipment ID or approval reference, and two vendors use recurring invoice numbers.
Question: Which response fits the analytic output?
- Record all fourteen pairs as duplicate payments.
- Discard the rule because a computer result cannot be audit evidence.
- Trace each pair to shipment, approval, and payment records, then classify the match as an exception or a valid repeat.
- Double the random sample before inspecting a flagged pair.
Answer and rationale: Choice 3 is strongest. This rule narrows attention; source records explain what the matching fields mean. Option two throws away useful prioritization. Option four adds volume while the known context gap remains. Record: the shipment or approval identifier that resolves one pair.
Case 3: Silver Creek Power and specialist scope
Fact pattern: In contrast, Silver Creek Power receives a valuation specialist’s estimate for an asset-retirement obligation. The report explains the discount rate and sensitivity range, but its procedures do not test whether the asset register is complete. Management asks the engagement team to use the midpoint.
Question: What next action preserves the evidence boundary?
- Use the midpoint because a specialist prepared the model.
- Reject the report without reading its methods or assumptions.
- Assess competence, scope, methods, and assumptions, then obtain evidence for the asset population and relevant assertion.
- Ask the specialist to approve the engagement conclusion.
Answer and rationale: Choice 3 is strongest. Technical work can support an engagement while completeness and responsibility remain with the engagement team. Option one treats an estimate as population evidence. That second choice discards useful support before testing its limits. Option four transfers a judgment that the team still owns. Record: the report scope, assumptions, asset register, and completeness check.
Read AUD sampling rationales as an evidence trail
After the attempt, review a missed case in this order: state the claim, identify the population or assertion, describe the selection or analytic signal, name the limitation, and choose the record that would close the gap. The sequence keeps the rationale attached to the fact instead of to an answer letter.
Next, write the strongest distractor in plain language. “The total agrees” is a reconciliation statement, not a completeness statement. “The rule flagged it” is a prioritization statement, not a duplicate-payment conclusion. “A specialist modeled it” is a technical-work statement, not proof that the asset list is complete. Your card should show exactly where the distractor overreaches.
Change one AUD sampling fact
For transfer, change one detail after the first attempt and predict the result before looking again. Add the two late batches to the sampling frame. Add a shipment ID to the analytic file. Expand the specialist’s scope to include the asset register. If the answer changes, name the new evidence that caused the change. If it does not, name the boundary that stayed controlling.
Choose one bounded repair
As a result, repair the earliest incomplete box. A missing population cutoff calls for a frame check. A missing analytic identifier calls for source-record tracing. An unclear specialist boundary calls for a scope and assertion review. Open one current source or exhibit, write the missing fact in one sentence, and retry one new case. More questions can wait until the card is complete.

Use the second visual as a decision record: map the signal, mark the boundary, write the next record, and compare the changed fact. The board is designed for a U.S. study session, where an evidence note should tell the next reviewer what remains unproven.
Make AUD sampling review repeatable
Finally, keep a single ledger with these fields: Blueprint row and checkpoint date; task verb; population and cutoff; selection method or analytic fields; exception or specialist limitation; response chosen; strongest distractor; person responsible for the remaining judgment; source checked; changed-fact result; and next action. Record the minutes used so speed can improve without erasing uncertainty.
Meanwhile, separate authority from practice. AICPA pages define the current tested boundary. NASBA pages handle jurisdiction and administration questions. The cases, explanations, evidence card, and visual boards here are independently written. Reopen the live sources before carrying a card into another testing window or a real engagement.
AUD sampling flashcards checklist
- Save one current Blueprint row and checkpoint date.
- Write the task verb before reading the options.
- Define the period population and cutoff.
- Describe what the selection or analytic signal can represent.
- Name the exception, assumption, or specialist scope limit.
- Identify who owns the remaining evaluation.
- Reject the strongest distractor in a complete sentence.
- Save one confirming record.
- Change one fact and explain whether the response changes.
- Schedule one bounded retry before adding volume.
Overall, strong AUD sampling flashcards leave an evidence trail. The goal is not to collect answer letters. It is to connect a fact to a boundary, a boundary to a record, and a record to the next focused action.
AUD sampling flashcards FAQ
Are these official exam questions?
No. The cases and rationales are original editorial practice. Use current AICPA material for authoritative scope and terminology.
What should I record after a missed case?
Record the task verb, population or scope boundary, strongest distractor, confirming record, and one changed-fact retry.
Should I add more items when time runs out?
Only after the first incomplete evidence box is repaired. More volume can preserve the same boundary error.
Does this method predict a passing score?
No. It creates observable study evidence and a focused next action; current rules, preparation, and performance determine the result.
Sources
- AICPA CPA Exam Blueprints
- AICPA AUD resources
- NASBA CPA Exam FAQ
- AUD review hub
