Audit Evidence & Opinions

AUD loves a “which evidence is most reliable?” question: rank the choices by source and form. The hierarchy runs auditor’s direct knowledge (inspection, recalculation, observation) > external confirmations and third-party documents > internally generated documents > inquiry alone. The classic tell is a choice that sounds thorough but rests on a client representation; per AU-C 500 / AS 1105, inquiry by itself is never sufficient and must be corroborated. Mind the modifiers: effective internal control makes internal evidence more reliable, and original documents beat photocopies or scans.

Don’t blur the related concepts. Assertions are what you’re proving (existence, completeness, valuation); audit evidence is what convinces you; analytical procedures and sampling are how you gather it. A frequent trap splits appropriateness into its two parts: relevance ties to a specific assertion (an A/R confirmation tests existence, not completeness), while reliability turns on source and nature. More of the same evidence can’t cure a relevance problem—piling on quantity won’t fix the wrong kind.

AUD loves to make you match the sampling method to the audit objective: attribute sampling tests controls (estimating a deviation/occurrence rate), while classical variables or PPS sampling tests dollar balances (estimating misstatement). The classic trap is the two-sided error pair. For controls, the risk of overreliance (assessing control risk too low) is the effectiveness killer the exam cares about, while underreliance only wastes work; for substantive tests, incorrect acceptance (concluding a materially misstated balance is fine) is the dangerous error, incorrect rejection merely costs extra testing. The acceptance/overreliance errors threaten the opinion.

A reliable tell: lower tolerable rate or tolerable misstatement, or higher desired confidence, drives sample size UP, and a larger expected error also raises it. Don’t confuse sampling risk (the sample isn’t representative) with nonsampling risk (wrong procedure or misread result) — only the former shrinks by enlarging the sample. Sampling is one technique for gathering evidence; it never substitutes for evidence that is sufficient and appropriate.

AUD loves to test the three phases: required as risk-assessment procedures during planning (technically driven by AU-C 315, often labeled “AU-C 520” on exams) and required in the final overall review under AU-C 520, but optional as substantive procedures. The classic MCQ “tell” gives a phase and asks whether AP is mandatory — pick “required” only for planning and final review; substantive AP is a choice you make when it’s more effective or efficient than tests of details. That question hinges on the precision of the expectation: it works best for stable, predictable relationships (e.g., recurring revenue), and you investigate any difference exceeding your pre-set threshold.

Don’t confuse AP with audit sampling or tests of details — AP studies relationships and ratios rather than individual items, so it never “vouches” a transaction. In a review engagement (governed by SSARS, not the audit standards), inquiry and AP are the primary procedures yielding limited assurance; in an audit they support the opinion alongside other evidence. Memory hook: “Plan and Pass” — AP bookends the audit, required at start and end, optional in the middle.

AUD rarely asks “what is an opinion”; it hands you a fact pattern and makes you pick the right type from a 2×2 grid. One axis is the problem: a GAAP departure (misstatement) versus a scope limitation (insufficient appropriate evidence). The other is severity: material but not pervasive versus material and pervasive. A known error points toward qualified/adverse; an evidence gap toward qualified/disclaimer. Under AU-C 700/705, a modified report leads with the renamed Opinion section, then a matching “Basis for Qualified/Adverse/Disclaimer Opinion” section (SAS 134 reordered these so Opinion comes first; some older banks still show the basis ahead of it).

The classic trap is mismatching the axes: students wrongly issue an adverse opinion (reserved for pervasive misstatements) when the auditor merely couldn’t gather evidence — pervasive that way is a disclaimer, while a contained problem on either axis is qualified (“except for”). Another miss: a going-concern doubt that is adequately disclosed keeps an unmodified opinion — it adds a separate going-concern section, not a modification. Hook: problem sets the row, pervasiveness sets the column.

On the exam, the tell is a fact pattern with no material misstatement and no scope limitation — the only cell in the opinion grid that yields a clean report. The trap is assuming that anything added to the report degrades it. It does not: an emphasis-of-matter or other-matter paragraph (AU-C 706), a justified change in accounting principle (AU-C 708), a going-concern section for substantial doubt (a separate headed section since SAS 134 — older banks still call it an emphasis-of-matter paragraph), or key/critical audit matters (KAM under AU-C 701; CAM for PCAOB issuers under AS 3101) all coexist with an unmodified opinion. The principle: extra communication ≠ modification.

Distinguish the neighbors: a qualified opinion adds “except for” for a material-but-not-pervasive problem; adverse (a pervasive GAAP departure) and disclaimer (a pervasive scope limitation) require pervasiveness. Item-writers exploit the standard-setter split — AICPA/AU-C for nonissuers, PCAOB for issuers — so read who’s being audited before picking the report.

AUD items hand you a fact pattern and force you onto the 2×2 grid (misstatement vs. scope limitation × material-but-not-pervasive vs. material-and-pervasive). The “tell” is a problem that is isolated to one account or disclosure — an unsupported inventory balance, one uncountable subsidiary, an omitted disclosure — rather than something fundamental to the statements as a whole. That confinement is what drives “qualified” over the more severe options; the answer hinges on pervasiveness, not dollar size (AU-C 705), a distinction that trips up candidates who fixate on the amount.

Watch two traps. A scope-limited qualified opinion is the less-severe twin of a disclaimer; an except-for misstatement is the less-severe twin of adverse. Distinguish a true qualification from an unmodified opinion with an emphasis-of-matter paragraph — emphasis-of-matter (e.g., a justified accounting-principle change, an unusually significant subsequent event) does not modify the opinion. Note that under current standards going concern gets its own separate report section, not an emphasis-of-matter paragraph (some older banks still file it under emphasis-of-matter). Memory hook: “except for” = surgical, one bad spot cut out; adverse/disclaimer = the whole patient is sick.

Master the auditor’s opinion as a 2×2 grid. One axis is the nature of the problem — a misstatement (the statements depart from the applicable framework, e.g., GAAP) versus a scope limitation (the auditor cannot obtain sufficient appropriate evidence). The other axis is severity — material but not pervasive versus material and pervasive. An adverse opinion sits in exactly one cell: a misstatement that is both material and pervasive.

These are the cells the exam makes you discriminate. A qualified (“except for”) opinion is a material-but-not-pervasive misstatement; a disclaimer of opinion is the evidence-axis twin of adverse — a scope limitation that is material and pervasive. “Pervasive” (AU-C 705) means the effect is not confined to specific elements, or — if confined — represents a substantial proportion of the statements, or is fundamental to users’ understanding. In the report itself, an adverse opinion is preceded by a “Basis for Adverse Opinion” section that describes the misstatement and, where practicable, quantifies its effect.

AUD item-writers run a two-by-two matrix: the nature of the problem (scope limitation versus GAAP misstatement) crossed with pervasiveness. A disclaimer lives in one cell — the pervasive extreme of a scope limitation (or, rarely, when multiple undetected uncertainties make an overall opinion impossible). The tell is language like “unable to obtain sufficient appropriate audit evidence,” management restricting access to records, or being engaged after year-end so you can’t observe inventory and no alternative procedures work. Per AU-C 705 (SAS 134’s reordered report), the Opinion section states the auditor does not express an opinion, the basis becomes “Basis for Disclaimer of Opinion,” and Key Audit Matters are prohibited.

The classic trap is confusing the axes. Compared with a qualified opinion (material but not pervasive — “except for”), a disclaimer is the pervasive extreme of missing evidence; compared with an adverse opinion, a known pervasive misstatement, a disclaimer is about missing evidence, not wrong numbers. Memory hook: “Can’t see it” = disclaimer (scope); “It’s wrong” = adverse (GAAP).