Brett LindenbergLearning assessment strategies are the planned ways you collect evidence of what learners know and can do. While people are still learning, use assessment for learning so you can adjust your teaching: diagnostic pre-tests, checks for understanding, retrieval quizzes, and self- and peer assessment. At the end, use assessment of learning to confirm what they can do: performance tasks, portfolios, competency checks, and final exams. Pick each strategy by the decision the evidence has to support. In the short demo below, choose what an instructor needs to know after a forklift safety lesson and see which assessment method fits that need.
Types of assessment strategies: for, as, and of learning
Assessment strategies fall into groups by purpose. In 2002, the Assessment Reform Group defined assessment for learning as seeking and interpreting evidence that learners and their teachers use to decide where the learners are in their learning, where they need to go, and how best to get there. Assessment of learning sums up achievement at the end. Assessment as learning uses the act of being assessed as practice: recalling an answer or judging your own work builds the learning itself.
| Type | Question it answers | When | Who uses the evidence | Examples |
|---|---|---|---|---|
| Diagnostic | What do learners already know? | Before instruction | Instructor, to plan | Pre-test, skills inventory, placement quiz |
| Formative (assessment for learning) | Is learning on track right now? | During instruction | Instructor and learner, to adjust | Checks for understanding, exit tickets, in-video questions |
| Assessment as learning | Can learners recall the material and judge their own work? | During instruction | Learner | Retrieval quizzes, self-assessment, confidence ratings |
| Summative (assessment of learning) | Did learners reach the standard? | End of a unit, course, or program | Instructor, school, or employer | Final exam, performance task, certification |
One task can serve more than one purpose. A quiz is formative when you use the results to change the next lesson, and summative when it decides a grade or a certification. For more on the formative side, read what formative assessment is and our list of formative assessment techniques.
Five assessment for learning strategies
Dylan Wiliam and Marnie Thompson grouped formative practice into five key strategies in a book chapter published in 2008. The FaSMEd toolkit hosted by Newcastle University lists them as follows. Each comes with an idea you can use in your next session.
- Clarify and share learning intentions and success criteria. Show learners a strong and a weak example of the same task, and ask which one meets the standard and why.
- Engineer discussions, questions, and tasks that reveal understanding. Before you move to the next idea, ask a hinge question: a multiple-choice item in which each wrong answer points to a known misconception.
- Give feedback that moves learners forward. Tell each learner what to do next, and give them time in the session to do it.
- Activate learners as instructional resources for one another. Have pairs check each other's work against the success criteria before anyone hands it in.
- Activate learners as owners of their own learning. End the session by asking learners to rate their confidence on each objective and name one thing they will review.
12 learning assessment strategies with examples
Each strategy below says what evidence it gives you, with examples from classrooms and workplace training. Where a strategy works inside a recorded lesson, we show how.
1. Diagnostic pre-assessment
A short test before instruction shows what learners already know, so you can skip what they have mastered and plan for the gaps. Keep it ungraded, and tell learners why you are asking. Example: before a spreadsheet course, give five tasks that range from sorting a column to writing a lookup formula. In a training video, an opening question can use a change time answer action to send learners who answer correctly past the introduction and the rest to the basics.
2. Checks for understanding during instruction
Quick, frequent checks tell you whether to move on or teach an idea again. Classroom versions include mini whiteboards, answer cards held up at the same time, and exit tickets. In workplace training, a poll in a live session or a question placed right after a key concept in a recorded lesson does the same job. The value comes from what you do with the result: if most learners miss a hinge question, teach the idea again a different way. In a Mindstamp video, a question pauses playback at the moment the concept is taught, and the question report shows how many learners chose each answer.
3. Retrieval practice quizzes
Low-stakes quizzes that make learners recall the material help them remember it longer. In a 2006 study by Roediger and Karpicke, students who read a passage once and then practiced recalling it three times remembered 61% of it a week later, compared with 40% for students who read it four times. A 2013 review by Dunlosky and colleagues rated practice testing and distributed practice as the two study techniques with high utility. Combine them: ask a few questions a day after the lesson, again a week later, and again a month later. Self-check quizzes with an answer key use the same principle, such as this five-question Islamic quiz with answers from the Jibreel app. For workplace methods, see our guide to retaining information better.
4. Self-assessment and confidence checks
Self-assessment asks learners to judge their own work or understanding against clear criteria. It builds the monitoring skill that the Education Endowment Foundation describes in its guidance on metacognition: checking your progress during a task, between planning the task and evaluating the result. A simple version is to ask learners how confident they are before they see whether an answer is right. A learner who is confident and wrong has a misconception to fix. A learner who is right but unsure needs more practice. In a video lesson, a rating question or a Likert scale question placed after a scored question captures both signals.
5. Peer assessment
Learners review each other's work against a rubric. In a 2020 meta-analysis by Double, McGrane, and Hopfenbeck of 54 controlled studies, peer assessment improved academic performance with an overall effect size of g = 0.31, a small to medium effect. Train learners on sample work before they review peers, and write each rubric level in plain terms. Example: sales reps swap recorded pitches and score each other on the discovery questions asked, the objections handled, and whether the call ended with an agreed next step. Our free rubric generator drafts criteria and level descriptions that you can edit.
6. Performance-based and authentic tasks
Performance tasks ask learners to do the real thing, or something close to it, and score both the process and the result with a rubric. Authentic assessment puts those tasks in realistic conditions. Examples include a nursing student who performs a medication check on a simulated patient, a new manager who leads a recorded feedback conversation, and a technician who finds a fault on a training rig. These tasks give stronger evidence of skill than recall questions, and they take longer to design and score. In Mindstamp, a video response question asks the learner to record up to 120 seconds on camera, such as a pitch or a walkthrough of a procedure. Video and voice answers are not scored automatically, so a reviewer watches each one in the report and scores it with the same rubric used on the job.
7. Scenario-based assessment
A scenario shows a realistic situation, asks the learner what to do, and then shows what happens next. It tests judgment, which recall questions rarely reach. Example: a support agent watches a call with an angry customer, chooses a response, and sees how the customer reacts, with each choice leading to a different next scene. In Mindstamp, video branching plays the consequence of each choice, and the report records the path each learner took. For design steps, see our guide to scenario-based training.
8. Portfolio assessment
A portfolio collects a learner's work over time, usually with a short reflection on each piece, to show growth. Writing courses use drafts next to final versions. Design courses use project portfolios. Workplace programs often use a log of completed tasks that a supervisor signs off. Agree in advance on what goes in, how often learners add to it, and the criteria for the final review. Without those rules, a portfolio becomes a folder of unrelated files that nobody can grade consistently.
9. Competency-based (mastery) assessment
Learners move on once they show mastery of a defined competency, however long that takes. Western Governors University runs its degrees this way: WGU says students progress through courses as soon as they can prove they have mastered the material. In workplace training, a competency check often pairs a knowledge test with an observed task. Forklift training is a regulated example: the OSHA standard 1910.178 requires formal instruction, practical training, and an evaluation of each operator's performance in the workplace, repeated at least once every three years. Set the passing standard before the first learner tries, and allow a retest after more practice. In a video lesson, Repeat Until Correct makes the learner try a question again until they answer it correctly, and every attempt stays in the report. See how passing scores and retries work.
10. Adaptive assessment
An adaptive test changes difficulty based on the learner's answers, so each learner spends most of the test on questions near their level. NWEA's MAP Growth adjusts after each question: a correct answer brings a harder question, and a wrong answer brings an easier one. The GRE General Test adapts by section: the second Verbal and Quantitative sections are harder or easier depending on how the test taker did on the first (checked October 2026). Item-by-item adaptive testing needs a large, calibrated question bank. A simpler version works in any course: route learners by their answers to a short diagnostic, as in strategy 1. For more models, see our adaptive learning examples.
11. Gamified assessment
Points, levels, timers, and leaderboards turn practice quizzes into games, a format that classroom tools such as Kahoot! made familiar. Use game mechanics for low-stakes retrieval practice, where speed and repetition help. Keep them out of high-stakes decisions, where a timer can end up measuring reading speed instead of knowledge. For workplace ideas, see gamification in corporate training.
12. Summative assessment
Summative assessment certifies achievement at the end of a unit, course, or program through final exams, capstone projects, certification tests, and observed sign-offs. Write it from the learning objectives, test each objective at the level it was taught (recall, application, or judgment), and set the pass mark before anyone takes it. Combine formats, such as a scenario-based knowledge test plus an observed task, so one weak format does not decide the result. In Mindstamp, a certificate can require a passing score and a watched percentage before the learner can download it. See completion certificates for the setup.
How to choose an assessment strategy
Start from the decision you need to make, then choose the strategy that gives evidence for it. A course usually needs at least one strategy during learning and one at the end.
| Decision you need to make | Strategy | Effort to build | Effort to score | Works in a video lesson |
|---|---|---|---|---|
| Where to start teaching | Diagnostic pre-assessment | Low | Low, scored automatically | Yes, with answer routing |
| Whether to teach an idea again now | Checks for understanding | Low | Low | Yes |
| Whether learners will remember | Retrieval practice quizzes | Low | Low | Yes |
| Where learners think they need help | Self-assessment and confidence checks | Low | None | Yes, with rating or Likert questions |
| Whether learners can judge quality | Peer assessment | Medium | Medium, done by peers | Partly: recorded work for peers to review |
| Whether learners can do the task | Performance-based and authentic tasks | High | High, with a rubric | Partly: video and voice responses |
| Whether learners make sound decisions | Scenario-based assessment | Medium to high | Low, path is recorded | Yes, with branching |
| How learners grew over time | Portfolio assessment | Medium | High | No |
| Whether a learner is ready to move on | Competency-based assessment | High | Medium | Partly: knowledge checks, not observed tasks |
| Where to place each learner | Adaptive assessment | High, needs a question bank | Low | Simple routing only |
| How to keep learners practicing | Gamified assessment | Medium | Low | Partly |
| Whether learners met the standard | Summative assessment | Medium | Medium | Yes, for knowledge and scenario items |
Assessment best practices
These practices apply to every strategy above, in a classroom or a corporate course.
- Align each assessment with an objective. Write the objective with a verb you can observe, such as "identifies," "calculates," or "demonstrates," and then write items that require that verb. A recall question cannot assess an objective that says "applies."
- Share the criteria in advance. Show the rubric and a sample of strong work before the task, so learners know the standard they are aiming for.
- Use more than one measure before a high-stakes decision. Pair a knowledge check with a performance task or an observation.
- Give feedback that says what to do next. Hattie and Timperley's 2007 review of feedback research frames useful feedback around three questions: Where am I going? How am I going? Where to next? Their review found feedback about the task and the process more effective than praise directed at the person.
- Keep most assessment low stakes and frequent. Learners show what they do not know when a wrong answer costs them little.
- Score consistently. Have raters score the same sample work together before they score alone, and spot-check a share of the scored work.
- Design for every learner. The CAST UDL Guidelines, updated to version 3.0 in July 2024, call for multiple means of engagement, representation, and action and expression. For assessment, let learners show the same skill in more than one way where the skill allows it, for example by speaking an explanation instead of writing it.
- Act on the results. Change the next lesson, teach the most-missed concept again, or rewrite a question that strong learners keep getting wrong.
How to build assessment into video lessons
Recorded lessons are where assessment most often goes missing. Learners press play, and the only record is that the video ran. These changes make a video lesson produce evidence:
- Put a question at the moment each key idea is taught, instead of saving every question for a quiz at the end. In Mindstamp, a question pauses the video, and you can turn off skipping on the questions that must be answered.
- Write feedback for each answer choice, and send a wrong answer back to the explanation with a change time action.
- Turn on Repeat Until Correct for practice questions. Every attempt is recorded, so a first miss still shows in the report.
- Use a video or voice response when the learner has to show or say the skill.
- Identify learners with lead capture, a personalized link, or a SCORM package launched from your LMS, so each answer belongs to a person.
- After the first group finishes, review the question report and fix the questions, or the parts of the lesson, that most learners miss.
Our free training assessment generator drafts questions, answers, and explanations from your training material, and Mindstamp can draft multiple-choice questions from a video's transcript. Review every drafted item before you publish it. To compare platforms, see our lists of video assessment tools and online assessment tools for teachers. To track attention alongside learning, read how to measure and improve student engagement.
Common questions
Which of the following best illustrates the idea of monitoring during learning?
Monitoring means checking your own understanding or progress while you work, then changing your approach when something is not working. A typical example is a learner who stops at the end of a section of a technical manual, finds that they cannot explain it in their own words, and rereads it or asks for help. The Education Endowment Foundation's metacognition guidance places monitoring between planning a task and evaluating the result. A confidence rating after each quiz answer is a simple way to build the habit.
Which of the following is considered a best practice for assessment?
Align each assessment with a stated learning objective, so that every item tests what the course set out to teach. Other sound practices are to share the criteria in advance, use more than one measure before a high-stakes decision, give feedback that says what to do next, and use the results to change instruction.
What is the difference between assessment and evaluation?
Assessment collects evidence about learners: what they know and can do. Evaluation judges the value of something larger, such as a course or a training program, and often uses assessment results as one input. For workplace programs, see our guides to training evaluation methods and how to measure training effectiveness. If you need to show a financial return, read how to calculate the ROI on training.
What are the newest assessment methods?
Recent methods include computer-adaptive tests, branching video scenarios, recorded video and voice responses, and AI tools that draft questions from course material. They change how you collect evidence. Alignment with objectives, clear criteria, and consistent scoring still decide whether that evidence means anything. Check every AI-drafted item for accuracy and for whether it tests the objective you intended.
What are effective assessment strategies for elearning?
In self-paced elearning, no instructor sees learners work, so build the evidence into the course: questions with feedback at each key concept, scenario decisions, a confidence rating, and a final task that mirrors the job. Make sure each response is tied to an identified learner and lands in a report that someone reviews.
Start with one lesson
Pick a lesson you teach often. Add a three-question diagnostic at the start, two checks for understanding where learners usually get confused, and one task at the end that asks learners to use the skill. After the first group finishes, find the question most learners missed and change that part of the lesson. For more ways to collect evidence during learning, see our roundup of formative assessment online tools.
