Formative vs summative assessment: differences, examples and when to use each
Formative vs summative assessment comes down to what you do with the result. A formative assessment checks learning while it is still happening, so the instructor and the learner can change course. A summative assessment measures learning at the end, against a standard, to support a decision such as pass, fail or certify.
Most pages on this topic are written for K-12 and college classrooms. This guide covers those basics, then adds what trainers and L&D teams need: workplace examples, a way to classify ambiguous formats such as quizzes and role plays, and a sample course plan. Every source is linked, and the research section explains where the sources differ.
Key takeaways
- Formative assessment guides the next step during learning, usually at low or no stakes. Summative assessment judges the outcome at the end, often at higher stakes.
- Format does not decide the label. A quiz, a role play or a project is formative or summative depending on the decision its result feeds.
- Research suggests formative assessment helps only when someone acts on the evidence, so build in time to use the results.
- A good course uses formative checks throughout, keeps their items separate from the final, and reserves one or two summative decisions for the end.
What is the difference between formative vs summative assessment?
Formative assessment monitors learning while it is in progress and gives feedback, usually at low or no stakes. Summative assessment evaluates learning at the end of a unit or course against a standard, usually at higher stakes. The comparison chart below sets the two side by side on the points that matter when you design a course.
Carnegie Mellon’s Eberly Center states the split plainly: its guidance on formative and summative assessment says the goal of formative assessment is to “monitor student learning to provide ongoing feedback,” while summative assessment aims to “evaluate student learning at the end of an instructional unit” by comparing it against a standard or benchmark.
| Point of comparison | Formative assessment | Summative assessment |
|---|---|---|
| Purpose | Monitor learning and guide the next step | Judge whether learning met a standard |
| Decision it feeds | What to reteach, practice or skip | Pass, fail, grade, certify or sign off |
| Timing | During instruction, often repeated | End of a unit, module or course |
| Stakes | Low or none | Often high |
| Feedback | Immediate and specific, ideally explaining why an answer is wrong | Often a score or outcome with limited detail |
| Typical formats | Knowledge checks, polls, practice quizzes, drafts, coached role plays | Final exams, graded projects, certification tests, observed sign-offs |
| Who acts on the result | The instructor and the learner | The employer, client, regulator or program owner |
| Main risk | Grading it too early, so learners hide gaps | It arrives too late to help anyone improve |
What does the research say about formative assessment?
Formative assessment improves learning only when someone acts on the evidence. Black and Wiliam’s 1998 review reported typical effect sizes between 0.4 and 0.7, while a 2022 meta-analysis of K-12 reading found a smaller effect of 0.19. Both point to the same condition: the result has to change what happens next.
In “Inside the Black Box”, Paul Black and Dylan Wiliam wrote that assessment “becomes formative assessment when the evidence is actually used to adapt the teaching to meet student needs.” They reported that “typical effect sizes of the formative assessment experiments were between 0.4 and 0.7,” larger than most educational interventions.
A later meta-analysis in Frontiers in Psychology by Xuan, Cheung and Sun pooled 48 studies and 116,051 K-12 students. It found a positive but modest effect on reading achievement (ES = +0.19). The two sources differ in scope, and the 2022 paper covers K-12 reading only, so treat both as evidence about the mechanism, feedback that gets used, and not as a forecast for your course.
There is also evidence on why frequent low-stakes checks help. In a 2006 study in Psychological Science, repeated studying beat repeated testing on a test given five minutes later. On delayed tests, though, “prior testing produced substantially greater retention than studying.” Expect the payoff from formative quizzes at the later check, not in the same session.
How do you tell whether an assessment is formative or summative?
Ask what happens to the result. If it mainly changes what the instructor teaches or what the learner studies next, it is formative. If it mainly decides a pass, a grade or a credential, it is summative. The format does not settle it: a quiz, a role play and a project can each serve either purpose.
Three questions make the call faster and expose assessments that are trying to do both jobs:
- Who acts on the result? Instructor and learner point to formative. An employer, client or regulator points to summative.
- Can the learner still change the outcome? Yes points to formative. No points to summative.
- Does the result carry a consequence outside the course? No points to formative. Yes points to summative.
If the answers split, you have a design problem to settle before launch, not after. Carnegie Mellon notes that summative results can still be used formatively, for example to guide later work. Treat that as a second use of the data, not a reason to blur the purpose.
The table below applies the test to common formats.
| Format | Formative when | Summative when |
|---|---|---|
| Quiz | It gives feedback and carries no points or only completion credit | Its score counts toward a pass mark or final grade |
| Role play or simulation | A coach stops, gives feedback and lets the learner retry | A rater scores one attempt against criteria for sign-off |
| Practice test | Learners review the missed objectives and study again | It is the real exam, with a recorded outcome |
| Observation checklist | The observer coaches during the session | The observer records a competent or not-yet-competent result |
| Pre-test | It decides what to emphasize in teaching | It is used to exempt learners from the course |
| Project or case study | A draft gets comments before the final | The final version is graded against a rubric |
Pro tip: Finish The Sentence First. Before launching any assessment, complete this sentence in its instructions: “This result will be used to ___.” If you cannot finish it, or two colleagues finish it differently, the assessment is not ready to run.
What are examples of formative and summative assessment?
In classrooms, formative examples include concept maps and early drafts, and summative examples include midterms and final projects. At work, formative examples include scenario checks during onboarding and coached role plays, while summative examples include certification exams, compliance tests and observed practical sign-offs.
What are classroom examples?
The Carnegie Mellon guidance lists specific cases. For formative assessment, it gives sketching a concept map in class, writing one or two sentences on the main point of a lecture, and submitting a research proposal for early feedback. For summative assessment, it gives midterm exams, final projects, papers and senior recitals.
What are workplace examples?
Workplace examples are scarce in the top results, so this table offers a set to adapt. They are editorial examples, not drawn from any study or customer.
| Training type | Formative check | Summative check |
|---|---|---|
| New-hire onboarding | Short scenario questions after each module, with explanations | End-of-program test on policies and tools |
| Compliance or safety | Ungraded practice questions on rules that learners often misread | Scored test with a documented pass mark |
| Customer service or sales | Coached role play with a retry | One scored call simulation rated against a rubric |
| Technical or software skills | Guided exercises with instant feedback | Timed task completed without help |
| Leadership and soft skills | Peer feedback on a draft action plan | Manager-rated behavior review weeks later |
| Certification course sold to clients | Practice exams with item-level feedback | Proctored or secured final exam |
The Kirkpatrick model helps place these in a training evaluation. Kirkpatrick Partners describes Level 2 as whether participants acquire the intended knowledge, skills, attitude, confidence and commitment, and Level 3 as whether they perform critical behaviors on the job. Formative checks and end-of-course tests both sit at Level 2, and a manager sign-off weeks later is a summative check at Level 3. That mapping is our editorial judgment, not part of Kirkpatrick’s published model.
How do you combine formative vs summative assessment in one course?
Use formative checks at every stage and keep one or two summative decisions for the end. A four-module course might open with an ungraded pre-check, add short checks after each module, and finish with one scored test and a later on-the-job sign-off. The plan below is illustrative, not a validated template.
| Stage | Type | Format (illustrative) | Counts toward result? | What happens to the result |
|---|---|---|---|---|
| Before module 1 | Formative | 8-item pre-check | No | Instructor decides what to emphasize |
| After each module | Formative | 3 items with explanation feedback | No, or completion only | Learner reviews the missed objective |
| Live session | Formative | Polls and a coached role play | No | Instructor adjusts pace and examples |
| End of course | Summative | 30-item scored test | Yes | Pass at 80% (placeholder) |
| 30 days later | Summative | Manager observation checklist | Yes | Sign off or plan a refresher |
The staging idea is developed in assessing blended learning before, during and after the session, and the pre-check works best when paired with a post-test, as shown in pre and post training assessments for instructor-led courses. The 80% pass mark is a placeholder: how cut scores are set for professional courses explains how to choose a defensible one.
For the 30-day step, observation checklists for practical training cover how to build criteria that different raters apply the same way.
Pro tip: Tag Items By Purpose. Mark each question as practice or final in your item bank, and draw the final only from items tagged final. If practice items reappear on the exam, scores partly reflect memory of the practice, not skill.
When does formative vs summative assessment go wrong?
The common failures are about purpose, not format: results nobody uses, labels that do not match the stakes, and practice items that leak into the final. Most are fixed by deciding the purpose before writing the assessment.
- Grading every check. Heavy points give learners a reason to hide what they do not know.
- Feedback that arrives too late. A formative result delivered after the next module cannot change anything.
- Ignoring the results. Black and Wiliam’s condition was that the evidence is used, so schedule time to reteach.
- Only a final test. You find the gaps when it is too late to close them.
- High-stakes decisions from low-stakes data. An unsupervised practice quiz is weak evidence for certifying someone as competent.
What should you do next?
Pick one course you run and list every assessment in it. Beside each, finish the sentence “This result will be used to ___” and mark it formative or summative. Then check that each module has a formative check before the summative one, and that no pass or certify decision rests only on practice data.
If you build these in an assessment platform, check that it can tag items and report by learner and group. SimpliAssess’s published features page lists item banking, tag management and analytics and reporting, and its FAQ says results can be exported to a spreadsheet. Its published pages do not describe a setting that labels an assessment formative or summative, so ask about that when you request a SimpliAssess demo.
Frequently asked questions
Is a quiz formative or summative?
It depends on the decision the score feeds. A short quiz that gives feedback and does not count toward a grade is formative. A scored quiz that contributes to a pass mark or final grade is summative. The format is the same either way, so state the purpose before learners take it and do not change it afterward.
Can an assessment be both formative and summative?
The results can serve both purposes. Carnegie Mellon’s Eberly Center notes that summative information can be used formatively when students or instructors use it to guide later work. To avoid confusion, design each assessment for one primary purpose and treat a second use, such as reviewing exam results to improve the course, as a bonus.
Should formative assessments be graded?
Usually not. Carnegie Mellon’s Eberly Center describes formative assessments as generally low stakes, with low or no point value. Heavy grading gives learners a reason to avoid showing what they do not know, which defeats the purpose. If you need learners to complete the checks, award credit for completion instead of for correct answers.
How many formative checks does a course need?
No source sets a required number. A practical rule is one short check per learning objective or module, placed before the summative test, with enough time for learners to act on the feedback. Count checks by objectives, not by minutes, and cut any that nobody uses. This is editorial guidance, not a published standard.
Which is more important, formative or summative assessment?
Neither replaces the other. Summative assessment supplies the evidence for a pass, grade or credential, and formative assessment supplies the evidence needed to improve results before that point. A course with only summative tests finds problems too late, and one with only formative checks cannot support a clear decision about competence.
Is a certification exam summative?
Yes, in almost every case. A certification exam evaluates what a candidate can do against a defined standard at the end of preparation and leads to a decision, which matches the Eberly Center’s description of summative assessment. The practice exams candidates take beforehand are usually formative, provided they give feedback and do not carry the credential.
