The word “summative” describes the purpose of the evidence, not a particular format. The same presentation might be formative when students receive feedback and revise it, then summative when the final version is graded. What matters is when the evidence is used and what decision it supports.
Summative and formative assessment
| Dimension | Summative | Formative |
|---|---|---|
| Primary purpose | Judge achievement after learning | Improve learning while it is still developing |
| Timing | End of a unit, course, or major stage | During instruction and practice |
| Typical consequence | Grade, credential decision, or documented achievement | Feedback, reteaching, revision, or adjusted practice |
| Examples | Final exam, capstone, final performance, portfolio | Draft review, practice quiz, hinge question, peer feedback |
| Student opportunity | Demonstrate integrated learning | Identify gaps and act on feedback |
Common types of summative assessment
- Selected-response exams, including well-designed multiple-choice questions.
- Constructed-response exams with short answers, problems, or essays.
- Projects that require students to apply knowledge to a defined problem.
- Performances, demonstrations, simulations, presentations, or oral examinations.
- Portfolios curated to demonstrate achievement across several outcomes.
- Capstone work that integrates learning across a program or course sequence.
In higher education, the strongest format is the one that elicits the evidence required by the learning outcome. Recall can be assessed efficiently with selected-response items. Evaluation, design, clinical reasoning, communication, and performance usually require richer evidence. Convenience matters, but it should not determine the construct being measured.
How to create a defensible summative assessment
- Clarify the decision. Identify exactly what the assessment result will mean and who will use it.
- Map each task or question to a learning outcome and the level of thinking it requires.
- Sample the domain broadly enough. One convenient task rarely represents an entire course outcome.
- Define criteria before grading. Use an answer key, scoring guide, or rubric appropriate to the evidence.
- Review accessibility, instructions, timing, permitted resources, and opportunities for clarification.
- Moderate the assessment after use: inspect ambiguous items, unexpected patterns, and grading inconsistencies before the next offering.
Benefits and limitations
- Benefits: documents achievement, creates a common decision point, supports credentials, and can integrate learning across a course.
- Limitations: arrives too late to repair some learning gaps, may overrepresent performance on one day, and can narrow learning when poorly aligned.
- Mitigation: combine multiple forms of evidence, prepare students through formative practice, publish criteria, and retain instructor review over consequential decisions.
A score is not automatically a valid summary of learning. Validity depends on alignment, sufficient evidence, appropriate conditions, and defensible interpretation. Reliability also matters: two comparable performances should not receive substantially different judgments because criteria were vague or grading drifted.