Ask how a course assessment is landing across cohorts and get a written read with an inline chart of pass rates, plus the questions people miss most. A snapshot on demand, grounded in your LMS data, so you know exactly which content to fix.
Analyze the results for [Data Handling & Privacy] across [our last three cohorts] and show me pass rates, average scores, and the questions people miss most.
An illustrative example. Your output will reflect your own sources, data, and prompts.
Reading the assessment results for Data Handling & Privacy across your last three cohorts, 214 learners in total. The overall pass rate is 78%, just under your 80% pass mark, and the miss pattern points at one module rather than one weak group, both shown below.
Across your last three cohorts of Data Handling & Privacy, 214 learners sat the assessment. The overall pass rate is 78% against an 80% pass mark, and the overall average score is 84%. So most people score well, but just enough are landing under the line to pull the pass rate below target.
The story is in the cohort split. Q1 cleared the bar at an 82% pass rate, but Q2 dipped to 74% — below the pass mark — and Q3 recovered only partway to 79%. Two of the three cohorts came in under 80%, which is why the blended number sits at 78%.
Pass rate by cohort (pass mark 80%)
A single cohort dipping could be noise. But when I look at which questions are dragging scores down, the same items miss across all three cohorts, so this reads as a content problem rather than one weak group of learners. Here are the four questions people get wrong most often:
| Question | Answered correctly |
|---|---|
| Q7 – Data retention periods | 46% |
| Q3 – Lawful basis for processing | 58% |
| Q11 – Breach notification timeline | 61% |
| Q5 – Classifying personal data | 69% |
Only 46% answer Q7 – Data retention periods correctly, and 58% get Q3 – Lawful basis for processing. Q11 – Breach notification timeline sits at 61% and Q5 – Classifying personal data at 69%. The important part: the two weakest questions, Q7 and Q3, both map to the same section of the course — Module 3, Handling & Retention. When two questions from one module are the two worst on the whole test, the gap is in how that module teaches the material, not in the people taking it.
My read: this is a targeted fix, not a reason to re-run the whole course. Revise Module 3 — the retention-period and lawful-basis explanations in particular — then re-test the next cohort on those items. If Q7 and Q3 climb, you have your answer. The other three cohorts scored well enough elsewhere that a full rebuild would be effort spent where the content is already working.
Want me to pull the wrong-answer distribution for Q7 to see which option people pick instead, or compare these pass rates by department to check whether any one team is dragging the average?
The Assessment Results Analysis recipe reads your LMS assessment results on demand and returns a written analysis with one inline chart: pass rate by cohort, the overall pass and average scores, and a table of the questions people miss most. It points you at the specific content to revise, produced when you ask rather than a board that runs on its own.
Give Joy read-only access to your LMS, or add an export of the assessment results, per-question responses, and cohort records to the Knowledge Center.
Name the course and the cohorts you want compared, and ask for pass rates, average scores, and the questions people miss most. Joy reads the results at ask time and computes the numbers.
Joy returns a short written read with an inline chart of pass rate by cohort and a table of the most-missed questions. The figures in the prose match the chart and table.
Copy the read and numbers into your course-revision notes or your update to the course owner. Joy drafts the analysis; you act on it. Nothing is exported or scheduled for you.
Save this ask as a custom command on the assistant your team already uses, so anyone can run it in one step.
One inline chart ranks each cohort against your pass mark, so a dip is impossible to miss.
A table of the questions people get wrong most, ranked by how few answer them correctly.
Groups weak questions by module so you can tell a content gap from a one-off cohort.
Follow up for a single question's wrong-answer split or a per-department view, and Joy re-reads.
Compare pass rates across teams to see whether a gap is content-wide or team-specific.
Pull the wrong-answer distribution for one item to see which distractor people fall for.
Track a course's pass rate across more cohorts to see whether a revision actually worked.
Roll the question-level scores up to module level to rank which sections need the most work.
No. This is a written analysis produced on demand with one inline chart, not a persistent board. We dropped "Dashboard" from the name on purpose. When you ask, Joy reads your assessment results and returns the read for the cohorts you named. Nothing refreshes on its own between asks.
Joy ranks questions by how few learners answer them correctly, then checks the course map to see whether the weak ones cluster in one module. In the example, Q7 at 46% and Q3 at 58% both sit in Module 3, which points to a content gap rather than a weak cohort.
Yes. Ask a follow-up and Joy re-reads the data to break pass rates down by department, so you can tell whether a low score is content-wide or concentrated in one team.
Always. The pass rates in the prose are the same ones in the inline chart (Q1 82%, Q2 74%, Q3 79%), and the percentages in the most-missed table are the ones the analysis discusses, so you can quote either without reconciling them.
No. Joy drafts the analysis in chat; you copy it into your revision notes or your message to the course owner and share it yourself. Joy doesn't send messages or schedule recurring reports on your behalf.
Join the waitlist and be first to try this workflow when JoySuite launches.