Here's a familiar problem. A student gets a 3 on every row of a four-point rubric. Their work is solidly proficient. The gradebook divides 12 by 16 and reports 75%, a C. Something has gone wrong.
Why straight percentages mislead
On most four-point rubrics:
- 4 = exceeds the standard
- 3 = meets the standard
- 2 = approaching
- 1 = beginning
The levels describe quality, not a fraction of the work completed. A 3 means "good." Divided by 4, it reads as "average." And a 1 (an honest attempt) becomes 25%, which pulls averages down much like a zero does; see minimum grading.
Use a conversion scale
Instead of dividing, convert the rubric score with a table. A common example:
| Rubric average | Percentage |
|---|---|
| 4.0 | 100 |
| 3.5 | 93 |
| 3.0 | 85 |
| 2.5 | 78 |
| 2.0 | 72 |
| 1.5 | 65 |
| 1.0 | 60 |
| 0 (no evidence) | 0 or incomplete |
The exact numbers vary by school. What matters is that "meets the standard" lands where your school expects proficient work to land, and that the scale is published.
Multi-criterion rubrics
- Score each criterion.
- Average them, weighting criteria if some matter more (for example, double the weight for reasoning).
- Convert the average with your table.
A student with 4, 3, 3, 2 averages 3.0 and gets 85% rather than 75%.
Setting it up in your LMS
Many LMSs let you attach a custom grading scheme or set point values per rubric level that aren't evenly spaced (for example, levels worth 25, 21, 18 and 15 points out of 25). In Google Classroom, rubrics use the points you assign to each level, so set those deliberately rather than 1-2-3-4.
Be consistent
Use the same conversion across all rubric-graded work in a course and share it with students at the start. If you're designing rubrics, see holistic vs analytic rubrics and how to write a rubric for AI grading.