CuraCourseWhat the Analytics tab shows
Analytics summarizes how a class performed on one lesson, code exercise, or exam. Use it after students submit to find questions that were too easy, too hard, or confusing, and to see which wrong answers attracted students.
Before you begin
- Open the lesson, code exercise, or exam in view mode (not the editor or the course dashboard).
- Analytics uses the attempt your score policy counts toward each student’s grade. When the policy uses Average, item results average all of that student’s attempts, and response breakdowns include every attempt. Students who have not submitted are not included.
- For group work, each group counts once in the statistics and the CSV.
Open Analytics
- Open the item and select Analytics (the chart icon) in the right sidebar.
- Choose All Learners or a cohort in the filter. The filter is shared with Lightning Grader, so switching it in one tab switches it in the other.
- Select Refresh after more students submit or after you finish grading.
Overall scores
The top section shows how many students have a graded submission, then:
- Mean and median score, as a percentage and in points.
- Std. deviation: how spread out the scores are.
- Range: the lowest and highest scores.
- A histogram in 10% bins. Hover over a bar to see how many students fall in it.
- Shape: whether scores are roughly symmetric, skewed left (most scores high), or skewed right (most scores low).
Totals match the gradebook, including score modifiers. With fewer than 10 graded students, treat the statistics as a rough guide.
Item statistics
Each graded activity is listed in page order with its activity number and title. Select a title to scroll to that activity in the lesson.
| Statistic | What it means |
|---|---|
| Correct | Students who earned full credit, as n/n and a percentage. This is the item’s difficulty. |
| Avg score | The average percentage of the item’s points earned, including partial credit. |
| Discrimination | The point-biserial correlation between the item score and each student’s total score without that item. Higher values mean stronger students did better on the item. |
| Upper–lower 27% | The full-credit rate of the top 27% of students minus that of the bottom 27%. |
Discrimination is labeled Very good (0.40 or higher), Good (0.30–0.39), Fair (0.20–0.29), or Poor (below 0.20). A negative value means weaker students did better than stronger ones. That often points to a wrong answer key, an ambiguous question, or a trick question, so review the item.
Items answered correctly by 90% or more are tagged Very easy; items answered correctly by fewer than 30% are tagged Very hard. Responses still waiting for manual grading are counted as awaiting grading and left out of that item’s statistics until you score them. With the Average strategy, a student counts as correct on an item only when every attempt earned full credit.
If you add an activity after some students submitted, its card says how many students submitted before it existed. Their work is left out of that activity’s statistics instead of being counted as wrong.
Distractor and response breakdown
Select Distractor analysis or Response breakdown under an item to see what students actually answered:
- Multiple choice and multiple select: how many students chose each option. The correct option is marked with a check. A warning icon marks a wrong option chosen at least as often as the correct one. Options nobody chose are noted, since a distractor no one picks isn’t doing any work.
- Fill in the blank: the correct rate for each blank and the most common wrong answers.
- Categorize: the correct rate for each item and the categories students put it in instead.
- Ordering and Parsons problems: how often each step was in the right position, and how often each distractor was used.
Download the raw data
Select Download CSV at the top of the panel to analyze the results in a spreadsheet, R, or Python. The file has one row per student per activity (one row per group per activity for group work), with these columns:
- Student, email, and ID. For group work: group name, members, and group ID.
- Submission time and number of attempts counted.
- Total score and total possible points.
- Activity number, title, and type.
- Points earned, points possible, and percentage.
- Whether the student earned full credit: Yes, No, Ungraded, or Not on submitted version (the activity was added after they submitted).
- The student’s response, for multiple choice, multiple select, fill in the blank, categorize, and ordering activities. With several attempts, each is labeled (Attempt 1, Attempt 2, …).
Text that starts like a spreadsheet formula (for example, =, +, -, or @) is prefixed with an apostrophe so spreadsheets show it as text instead of running it.
The CSV respects the cohort filter, and the cohort name is added to the file name.
Common gotchas
- Statistics use the current published version of the page.
- Activities worth 0 points are not listed.
- If you edit an answer key, regrade the lesson so the statistics reflect the new key.