Is Using Test Banks Cheating? (And Do Professors Use Them?)

Reviewed and updated August 2026. Research findings below are from Cheng and Crumbley, “Student and professor use of publisher test banks and implications for fair play”, Journal of Accounting Education, as summarised by the Journal of Accountancy. Institutional rules vary — your syllabus and honour code govern your course.

Most confusion about test banks comes from the term itself. Students, professors and study sites all use it to mean different things, and the answer to “is this cheating?” changes completely depending on which one is in front of you.

What people actually mean by ‘test bank’

Four distinct things travel under this name. Sorting out which one you have is the whole question.

The middle row is where nearly everyone gets caught out. A leaked publisher bank sitting on a study site looks exactly like ordinary revision content — same interface, same subscription, sitting beside legitimate notes. It is still the instructor-only file, and finding it through a search engine does not convert it into a study resource.

So is using a test bank cheating?

When it is clearly a violation

You obtained material licensed to instructors, and you used it to prepare for an assessment that draws on it. That is the core case, and most institutions treat it as academic dishonesty whether or not the syllabus names test banks specifically, because it falls under obtaining an unfair advantage or using unauthorised materials.

It is worse, not better, when the questions turn out to be the actual exam. At that point the assessment measures who found the file rather than who understood the material, which is precisely the harm the rules exist to prevent.

When it is clearly fine

If your professor posted last year’s midterm, if the department maintains a practice-question archive, or if the textbook came with a student study guide, those are intended study resources. Using them is not cheating and you do not need to feel uneasy about it. Publishers produce student-facing practice materials on purpose, and they are separate products from the instructor test bank.

The grey zone, and how to resolve it

The honest grey area is material of uncertain origin: a shared drive from a student society, a folder passed down a year group, a set of questions a friend swears the professor released. The practical test is provenance. If you cannot say where it came from and whether it was released deliberately, treat it as restricted until you have asked.

Asking costs one email and permanently resolves the question. “Is it okay to revise from the 2024 midterm someone shared with me?” is a reasonable thing to send an instructor, and the answer protects you either way.

Your syllabus decides, not the internet

Rules vary genuinely between courses, and a professor can permit in one class exactly what another forbids. The documents that govern you are your syllabus, your course’s assessment instructions and your institution’s academic integrity code — in that order of specificity. No general article, including this one, overrides them.

Can you actually get caught?

Yes, and the method is more interesting than the usual claims about identity checks. The strongest published work on this comes from accounting education, where researchers built and validated a detection technique.

The stale-answer trick

Cheng and Crumbley, writing in the Journal of Accounting Education, developed a forensic approach that works like this: the instructor takes a publisher bank question and keeps the recognisable textual cues — the names, the scenario, the distinctive numbers a memoriser would latch onto. Then they change the correct answer, and keep the original, now-incorrect answer sitting in the list of options.

A student who understands the material works the problem and gets the new answer. A student who memorised the bank recognises the cue and selects the old answer — which is now wrong. Do that across a paper and the pattern is unmistakable, because getting a specific set of altered questions wrong in a specific way is not something understanding produces.

The elegance of it is that it needs no monitoring software, no proctoring and no accusation of possessing a file. The evidence is generated by the exam itself.

Performance that does not match your other work

The cruder signal is inconsistency. A student who has been mid-range all semester and then answers bank-derived multiple-choice questions near-perfectly while performing normally on essays and open-ended items invites a closer look. On its own that is a reason to ask questions rather than a finding — students do improve, and improvement is not evidence of anything. Combined with the stale-answer pattern above, it becomes considerably harder to explain.

What the research actually found

The Cheng and Crumbley study is the number everyone cites, and it is worth stating accurately, because it is routinely repeated in a broader form than it supports.

Studying accounting students, the authors found that 48% used publisher test banks, and that use produced a performance advantage of roughly 30% on exams. Their conclusion was about assessment validity rather than morality: when that many students have the questions, a professor building exams from an unmodified bank can no longer reliably distinguish students who understand the material from students who memorised it.

Two honest caveats. This was accounting coursework, not a survey of all students everywhere, so the 48% should not be quoted as a universal rate — disciplines with fewer standardised multiple-choice items will differ. And the finding describes an average advantage on bank-derived questions, not a guarantee of a better grade in a course.

Do professors use test banks?

Yes, widely, and the reasons are practical rather than lazy.

• Time. Writing good multiple-choice items is slow, skilled work, and it competes with teaching, grading and research.
• Expertise. Bank questions are written by subject specialists, frequently the textbook authors themselves, and are edited for the exact material the course covers.
• Scale. A large bank supports randomised versions, make-up exams and multiple sections without repeating items.
• Coverage. Banks map to chapters, which makes it straightforward to assess the syllabus evenly rather than over-testing whatever is easiest to write about.

There is also a real constraint in the other direction: some subjects simply have a limited pool of sensible questions. Where only so many distinct problems exist, items inevitably recur across semesters, which is exactly why some departments do not return exam papers.

What professors do to stay ahead of leaked banks

Instructors are generally aware their questions circulate, and published guidance for accounting faculty recommends a fairly consistent set of countermeasures:

• Modify bank questions so the correct answers differ from any circulating version
• Use bank items as a starting point and rewrite rather than deploying them unchanged
• Draw questions from more than one publisher
• Shift weight toward essay and open-ended questions that require demonstrated reasoning
• Discuss acceptable and unacceptable test-bank use with students early in the semester, rather than after an incident

That last one matters more than it sounds. A meaningful share of test-bank cases involve students who genuinely did not know the material was restricted, and courses that set the expectation explicitly at the start have less of that problem.

Test banks, old exams and study sites are not the same question

These get run together constantly, and they have different answers. If your situation is really one of the neighbouring ones, start there instead:

• Is using old exams cheating? — past papers, and where the line falls
• Do professors keep old papers? — why some exams are never returned
• Is Quizlet cheating? — including uploaded exam sets
• Is Course Hero cheating, and can you get caught?
• Is Chegg worth it? — and what Chegg shares with schools

Frequently asked questions

Is using a test bank cheating?

It depends which kind you mean. A publisher test bank is licensed to instructors only, and obtaining a copy is treated as academic dishonesty at most institutions. Practice questions your professor distributed, or a department’s official exam archive, are not cheating at all. The grey zone is a leaked publisher bank re-uploaded to a study site, which is still the restricted material even though it looks like ordinary revision content. Your syllabus and honour code decide the question, not the website hosting it.

Can you get caught using a test bank?

Yes, and not the way most students expect. Cheng and Crumbley, publishing in the Journal of Accounting Education, validated a detection technique in which instructors alter test bank questions while keeping the recognisable textual cues, change the correct numerical answer, and keep the original answer as one of the options. A student who memorised the bank picks the now-incorrect original answer, which is a pattern that stands out clearly across a paper.

Where do professors get their test questions from?

Textbook publishers supply copyrighted question banks to instructors who adopt their textbook, usually behind an instructor-verification login. Professors also write their own questions, adapt bank questions, and draw on department archives or shared collections. Many deliberately modify bank items precisely because they assume the originals have leaked.

Is it cheating to study from old exams?

Not if the professor or department distributed them, which is a common and legitimate study resource. It becomes a problem when the exam was never released, when you obtained it from a student who was told not to keep it, or when your course explicitly bars past papers. Because the same exam can be legitimate in one course and prohibited in another, the syllabus is the deciding document.

Do professors know their test bank questions are online?

Many do, and it shapes how they write assessments. Published guidance for instructors recommends modifying bank questions so the answers differ from circulating versions, pulling from multiple publishers, using bank items only as a starting point, and leaning on essay and open-ended questions. If a course still uses unmodified bank questions, that is usually a workload decision rather than ignorance.