Give Joy a question and the correct answer, and it drafts plausible wrong options, each one tied to a real misconception, so the item actually separates people who know the material from people who are guessing.
For this question — ["How long must we retain customer transaction records?"] with the correct answer ["7 years"] — write [4] options with plausible distractors.
An illustrative example. Your output will reflect your own sources, data, and prompts.
I drafted four options against your Records Retention Policy. One is clearly correct; the other three are each wrong for a reason a real learner could believe, so the item separates people who know the rule from people guessing.
Here are the four options. The three wrong ones each target a specific misconception, and every option reads as a number someone could genuinely defend, so nobody can find the answer by elimination alone.
| Option | Correct? | Why a learner might pick it (the misconception it targets) |
|---|---|---|
| 7 years | Correct | The retention period set in your Records Retention Policy for customer transaction records. |
| 3 years | No | Confuses this with the general business-record default many teams apply to routine paperwork, so it feels like a safe, familiar number. |
| 10 years | No | Over-cautious. Learners who half-remember tax or audit retention rules assume longer is always safer and round up. |
| Indefinitely | No | Sounds responsible — keep everything just in case — but it ignores data minimization, which requires you to stop holding data once its purpose ends. |
What makes these fair distractors is that each is plausible to someone who studied but missed one point. They are close to the right answer in length and grammar, so nobody can guess by picking the longest or the odd one out, and only one option holds up against the policy.
A few principles this item follows.
The same approach works on a policy figure. For "Under our expense policy, what is the per-meal limit for domestic travel?" with the correct answer $40:
| Option | Correct? | Why a learner might pick it (the misconception it targets) |
|---|---|---|
| $40 | Correct | The domestic per-meal cap stated in the expense policy. |
| $25 | No | Confuses the meal cap with the receipt-required threshold, another dollar figure in the same policy. |
| $50 | No | Confuses the domestic limit with the higher international per-meal allowance. |
| No fixed limit if reasonable | No | Sounds flexible and generous, but the policy sets a specific cap rather than a judgment call. |
Both items work the same way: the wrong answers are the mistakes your audience actually makes, so the score tells you who understood the rule and who reached for a plausible near-miss.
Want me to add a feedback line for each option that explains why it's right or wrong, or draft three more items at the same difficulty on this policy?
Distractor Generator takes a question and its correct answer and drafts options that a learner could genuinely land on for a specific, nameable reason. Each distractor targets a common error — a rule confused with a similar one, an over-cautious guess, a plausible-sounding shortcut — so choosing it reveals a real gap in understanding.
Paste the question stem and mark the right answer. Point Joy at the source policy or course content in the Knowledge Center so the distractors stay grounded in what you actually taught.
Tell Joy how many options you want and the difficulty. Joy drafts the wrong ones, each aimed at a misconception a real learner in your audience might hold.
Joy shows what each distractor targets, so you can check it is believable and fair. Ask it to swap one that is too easy, tighten wording, or make an option more tempting.
Copy the finished item into your LMS, a quiz form, or a doc. Drop it into your item bank alongside the others.
Save this ask as a custom command on the assistant your team already uses, so anyone can run it in one step.
Every wrong option maps to a specific error a real learner might make, so a wrong pick tells you something.
Options stay close in length, tense, and grammar so no answer gives itself away by its shape.
Distractors are wrong on the merits, not tricky-by-grammar, and only one option holds up against your source.
Distractors draw on your own policy or course material, so the near-misses reflect how people actually get it wrong.
Give Joy several questions at once and get a full set of options for each in one pass.
Add a short rationale for every option so learners see why their pick was right or wrong.
Ask for easier or harder distractors to match the level of the cohort or the exam blueprint.
Turn a plain recall question into a short scenario and get distractors that test judgment, not just memory.
A distractor is one of the incorrect options in a multiple-choice question. A good distractor is plausible enough that a learner with a specific misconception would choose it, which is what lets the question tell strong understanding apart from guessing.
A fair distractor is wrong on the merits, not because of a grammar trap or a hidden qualifier. It stays parallel to the other options in length and phrasing, targets a real misconception, and leaves exactly one option that is clearly best against the source material.
From your own content and the common errors your learners make. Point Joy at the policy or course material in the Knowledge Center and give it the correct answer, and it drafts wrong options grounded in how people actually get the topic wrong. You review and adjust before using them.
Usually no. As filler they weaken an item, because they reward test-taking tactics instead of knowledge. Joy avoids them by default and only suggests them when they genuinely test something. You can always ask for a different mix.
Yes. Tell Joy how many options you want and the difficulty, and it drafts to that spec. You can ask it to make a distractor more tempting, swap one that is too easy, or align the difficulty to an exam blueprint.
Join the waitlist and be first to try this workflow when JoySuite launches.