Definition:
A raw score is the straightforward, untransformed total number of points earned on a test — typically the count of correct answers. On a 50-item test where you answer 35 correctly, your raw score is 35. Raw scores are the starting point for all scoring systems but are rarely the final reported score on standardized tests, which convert raw scores into standard scores or scaled scores.
In-Depth Explanation
Calculation:
Raw Score = Total correct answers (or total points earned)
For tests with different point values per item: Raw Score = Σ (points earned per item)
Why raw scores aren’t used alone on standardized tests:
Raw scores have significant limitations for comparison:
| Problem | Example |
|---|---|
| Test form difficulty varies | Getting 35/50 on a hard version ≠ 35/50 on an easy version |
| No common scale | 35/50 on one test can’t be compared to 42/60 on another |
| No meaning without context | “35” tells you nothing without knowing the total, the difficulty, or how others performed |
Raw-to-scaled score conversion:
Standardized tests like JLPT, TOEFL, and TOEIC convert raw scores to scaled scores (or standard scores) using Item Response Theory or equating methods. This ensures that a scaled score of 100 means the same thing regardless of which test form was administered.
For the JLPT:
- The raw score (number correct) is not directly reported
- A scaled score (0–60 or 0–180 depending on section/total) is calculated using IRT
- This means two test-takers who both got 30 raw questions right on different test forms would receive the same scaled score if the IRT model determined those performances were equivalent
When raw scores are useful:
- Classroom tests where all students take the same test
- Formative assessments where the teacher knows the items
- Practice tests where the goal is tracking personal progress over time
- As input to more sophisticated scoring models
Related Terms
See Also
Research
- Crocker, L., & Algina, J. (2008). Introduction to Classical and Modern Test Theory. Cengage Learning. — Covers the relationship between raw scores and derived scores.
- Brown, J. D. (2005). Testing in Language Programs: A Comprehensive Guide to English Language Assessment. McGraw-Hill. — Practical treatment of score types in language testing.