Copying answers is the oldest problem in homework, and dynamic parameters in Pear Assessment attack it at the root. Instead of writing thirty versions of the same problem by hand, you write one question with variables in it — and the platform generates a different version, with different numbers, for every student who opens it. Same skill, same difficulty, no two papers alike.
This guide explains how questions with dynamic parameters work, how to build them, where they shine, and where their limits are — so you can decide which parts of your question bank deserve the treatment.
can pear assessment give every student a different version of the same question
Yes — that is precisely what dynamic parameters are for. When a question contains variables, Pear Assessment substitutes fresh values each time the question is delivered, so every student receives their own instance of the item. Maria might solve 34 × 6 while Jamal solves 47 × 3; both are answering your question about two-digit by one-digit multiplication, but copying a neighbor’s answer becomes useless because the numbers differ.
This works because the question is stored as a template rather than a fixed item. The text, the correct answer, and any worked feedback are all expressed in terms of the variables, and the system computes each student’s version — and its answer key — on the fly. Your reports still roll everyone’s results up under the one question you wrote, so you can see how the class handled the skill as a whole, exactly as you would with a static item.
Dynamic parameters are a premium capability in Pear Assessment, and they pair naturally with the other anti-copying settings covered in our guide to Pear Assessment kiosk mode and test security — randomization changes the numbers, while delivery settings protect the session itself.
What a dynamic parameter actually is
A dynamic parameter is a variable you declare inside a question. In the question editor you mark values with a variable name (written with an @ symbol, such as @a or @length), and then define how that variable gets its values. There are three basic ways to define one:
- A range — any value between a minimum and maximum, with a step you choose (for example, whole numbers from 5 to 60 in steps of 5).
- A set — a specific list of allowed values (for example, the numbers 3, 6, 9, 12 for a times-tables question, or a list of vocabulary contexts).
- A formula — a derived value computed from other variables, so the answer itself can be dynamic: if @a and @b are the sides, the correct answer might be defined as @a × @b.
The answer field uses the same variables, which is what makes auto-grading possible: the system knows each student’s numbers, so it knows each student’s correct answer. For numeric answers you can also think in terms of tolerance — a small accepted margin — where the subject allows approximate answers, though exact computed answers are the norm in math items.
Building a dynamic question, step by step
The build follows the same rhythm as any question in the item bank, with one extra layer. Start by writing the question in ordinary words with fixed numbers, and solve it yourself. This fixed version is your prototype: if the item does not work with one set of numbers, it will not work with five hundred.
Next, decide which values may vary and which must stay fixed. In a problem about buying @a notebooks at @price dollars each, both numbers can vary; the structure (multiplication, money context) stays. Replace the chosen numbers with variables everywhere they appear — in the stem, in any hints, and crucially in the answer.
Then define each variable: give it a range or set that keeps the problem sensible. This is the craft step. A range of 2–9 for a multiplicand keeps mental math friendly; a range up to 10,000 changes the skill entirely. Constrain decimal places where money or measurement is involved so students are not asked for absurd precision, and avoid combinations that produce negative or fractional results unless the question intends them.
Finally, generate several preview versions — not one. Pull up five or six instances and solve each. You are hunting for edge cases: the version where both variables are 1 and the question becomes trivial, the version where a division does not come out whole, the version where the “random” story stops making sense. Adjust the ranges until every preview is a fair instance of the skill. Our guide to the wider question-authoring toolkit, including the technology-enhanced types in Pear Assessment question types, will help you decide which items in your bank are worth parameterizing first.
A worked example
Take a static question: “A baker packs 24 muffins into boxes of 6. How many boxes does she fill?” As a dynamic item, it becomes: “A baker packs @total muffins into boxes of @per. How many boxes does she fill?” — with @per drawn from {4, 6, 8} and @total defined as a multiple of @per within a sensible span (say 24–96), and the answer computed as @total ÷ @per.
Every student still practices division with remainders avoided by design, but the numbers differ, the answer key differs, and a photographed answer key circulating in a group chat is suddenly worthless. Notice, too, what did not change: the context, the wording, the skill, and the marking scheme are identical for everyone — fairness is preserved while the surface varies. The same trick serves science (varying masses and forces in a friction calculation), economics (varying prices and quantities), and even language arts at the set level — swapping which excerpt, name, or setting from a defined list appears in a prompt.
Where dynamic questions earn their keep — and where they do not
The strongest use cases share one feature: repetition. Homework and practice sets, where students may collaborate a little too closely, become individually honest. Retakes and remediation improve, too — a student who failed the first attempt can receive genuinely fresh numbers rather than memorizing the first test’s answers, and you can assign the same template again next month as a fluency check without it feeling like a rerun.
There are honest limits. Constructed-response and essay questions cannot be parameterized — there is no computed key. Questions whose difficulty depends on carefully chosen numbers (a division that “just works” because the numbers were picked to be kind) can become unfair when randomized carelessly; the preview-and-constrain step exists for exactly this reason. And there is an authoring cost: a good dynamic item takes noticeably longer to build and verify than a static one. Spend that time on high-reuse items — the homework staples and retake templates — rather than one-off quiz questions.
It is also worth knowing that dynamic questions sit among the platform’s advanced capabilities: our feature-by-feature tour of Pear Assessment premium features places them in context alongside read-aloud, co-teaching, and rubric tools, so you can judge the upgrade case as a whole.
The official documentation walks the same build with field-level detail in its page on creating questions with dynamic parameters — a good companion when you build your first few items.
Mistakes that quietly break dynamic questions
Most failed dynamic items fail in predictable ways, and knowing the list saves debugging time.
The forgotten instance. The variable appears in the stem but not in a hint, a units label, or a follow-up part — so some students see a hint computed for numbers they do not have. Search the whole item for the original fixed numbers before you save.
The runaway range. A range chosen for variety rather than for the skill: allowing two-digit × two-digit products when the lesson was two-digit × one-digit. Every version is “valid,” but half the class is now being tested on next month’s skill. Ranges define the construct — set them from the lesson objective, not from what looks interesting.
The impossible combination. Individual ranges look fine, but together they can collide: a discount percentage above 100, a triangle whose sides cannot form a triangle, a remainder problem where the divisor exceeds the dividend. Cross-variable collisions are exactly what formula constraints and multi-version previews exist to catch.
The answer that is not unique. Especially in word problems, randomized values can accidentally create a second defensible answer — a “best buy” comparison where two options tie. Preview with fresh eyes, and where ties are possible, add a tie-breaker to the question (“if equal, choose the first”) or narrow the sets.
Troubleshooting a version a student flags
Sooner or later a student will say “my question was wrong,” and occasionally they will be right. Handle it by the book: open the student’s attempt, look at the exact version they were served, and solve it yourself from scratch. If the instance is broken — an impossible combination or a miscomputed key — the fair move is to void or credit that item for the class, fix the template, and note what changed in the item’s title or your own records so the bug does not silently return next term.
Keep a short personal log of which templates have run clean across a full class. That log becomes your trusted homework library — the items you can assign on a busy Friday with total confidence.
FAQ
Do students know their questions are different?
Not unless you tell them — each student simply sees a normal question. Many teachers do tell the class, because “everyone’s numbers are different” is itself a deterrent to copying.
Can I see which version a particular student got?
Reports are organized around the question you authored, and when you review a student’s attempt you see the version they were served — which is what makes discussing their specific work possible.
Do dynamic parameters work in every question type?
They fit best wherever answers are computed or selected from computed values — numeric and many multiple-choice items. Open-ended writing cannot be auto-generated this way.
How many variables should one question use?
Two or three cover the large majority of useful items. Every added variable multiplies the combinations you ought to preview, and long variable lists make the stem hard to read. If a draft needs five variables, it is probably two questions wearing a trench coat — split it.
Conclusion
Questions with dynamic parameters turn one well-built item into a fair test for every student, every time. Define variables with ranges, sets, or formulas, preview ruthlessly for edge cases, and reserve the effort for items you will reuse. Combined with sensible delivery settings, they are among the most practical integrity tools a digital assessment platform offers any busy classroom teacher.