forumguru.id Assessment is one of the most important components of the learning process. In a deep learning approach, assessment is not merely used to determine whether students have achieved a particular score. More importantly, assessment provides meaningful information about how students learn, how deeply they understand concepts, how they apply knowledge, and how they reflect on their learning experiences.
The implementation of deep learning in education emphasizes learning that is mindful, meaningful, and joyful. Therefore, assessment should also be designed to capture students’ learning processes and the development of their competencies, rather than focusing only on the final answer or test score.
In the Indonesian education context, the official framework for Pembelajaran Mendalam emphasizes three complementary functions of assessment: Assessment as Learning, Assessment for Learning, and Assessment of Learning. These functions allow teachers to obtain a more comprehensive picture of students’ learning progress and outcomes.
According to the 2025 Panduan Pembelajaran dan Asesmen, learning and assessment are viewed as an integrated cycle. Assessment provides information for designing learning and is also used to evaluate the effectiveness of learning that has taken place. Formative assessment is therefore given an important role in supporting students’ competency development.
What Is Assessment in Deep Learning?
Assessment in deep learning is a systematic process of collecting, interpreting, and using evidence of student learning to understand learning progress and determine appropriate next steps.
The purpose is not simply to answer the question, “What score did the student get?”
Instead, deep learning assessment asks broader questions:
- What does the student understand?
- Can the student explain a concept in their own words?
- Can the student apply knowledge to a real-life situation?
- Can the student analyze and solve problems?
- Can the student create something meaningful?
- Can the student collaborate and communicate effectively?
- Can the student reflect on their own learning?
- What support does the student need to improve?
This approach is consistent with the broader principle that effective assessment should support student learning and provide information that teachers can use to adapt instruction.
The OECD also emphasizes that formative assessment is valuable when information about student learning is used to guide feedback, teaching decisions, and students’ subsequent learning.
Why Is Assessment Important in Deep Learning?
Deep learning encourages students to go beyond memorizing information. Students are expected to understand concepts, connect ideas, apply knowledge, think critically, solve problems, communicate, collaborate, and reflect.
Consequently, conventional tests alone may not provide sufficient evidence of learning.
For example, a student may obtain a high score on a multiple-choice test but have difficulty explaining how a concept can be applied in real life. Another student may initially perform poorly on a written test but demonstrate excellent problem-solving skills through a project.
This is why assessment in deep learning should use multiple sources of evidence.
A balanced assessment system can help teachers:
- Identify students’ prior knowledge.
- Monitor learning progress.
- Provide meaningful feedback.
- Identify learning difficulties.
- Encourage student reflection.
- Measure achievement of learning objectives.
- Improve teaching strategies.
- Document students’ competencies and learning outcomes.
The goal is to obtain a more complete picture of student learning rather than relying on a single assessment technique.
Three Assessment Functions in Deep Learning
The Indonesian framework for Pembelajaran Mendalam identifies three important assessment functions: Assessment as Learning, Assessment for Learning, and Assessment of Learning.
These three functions should not be viewed as separate activities. They can complement one another throughout the learning process.
1. Assessment as Learning
Assessment as Learning positions students as active participants in the assessment process.
Students are encouraged to monitor, evaluate, and reflect on their own learning. This approach develops self-awareness, metacognition, and responsibility for learning.
Examples include:
- Self-assessment
- Reflective journals
- Learning logs
- Student checklists
- Learning portfolios
- Personal learning goals
- Reflection questions
- Peer assessment
For example, after completing a science project, students can answer questions such as:
What did I learn from this project?
What part of the project was difficult for me?
What strategy helped me solve the problem?
What would I do differently next time?
These questions encourage students to understand not only what they learned, but also how they learned.
2. Assessment for Learning
Assessment for Learning is primarily used to improve the learning process while it is still taking place.
Teachers collect evidence of student understanding and use that evidence to determine what should happen next.
Examples include:
- Diagnostic questions
- Classroom observation
- Concept maps
- Exit tickets
- Question-and-answer activities
- Formative quizzes
- Discussions
- Learning conferences
- Draft reviews
- Teacher feedback
For example, if a teacher discovers from an exit ticket that many students misunderstand a mathematical concept, the teacher can adjust the next lesson rather than simply continuing with the original plan.
This makes assessment an active part of teaching.
The OECD describes formative assessment as an ongoing process involving understanding student progress, providing feedback, and adapting teaching to students’ needs.
3. Assessment of Learning
Assessment of Learning is used to determine the achievement of learning objectives, usually at the end of a learning period or learning unit.
Examples include:
- Written tests
- Oral examinations
- Performance assessments
- Projects
- Products
- Reports
- Portfolios
- Presentations
- Summative assessments
However, assessment of learning in deep learning should not be limited to traditional examinations.
If the learning objective requires students to solve authentic problems, a project or performance assessment may provide stronger evidence than a conventional written test.
Models of Assessment for Deep Learning
Teachers can combine several assessment models according to learning objectives, student characteristics, and the learning activities being implemented.
1. Formative Assessment Model
Formative assessment is conducted throughout the learning process.
Its main purpose is to identify students’ current understanding and provide information that can improve subsequent learning.
A simple formative assessment cycle can be described as:
Learning Objective → Evidence → Interpretation → Feedback → Improvement
For example:
- The teacher establishes a learning objective.
- Students complete a short task.
- The teacher examines the evidence.
- The teacher identifies misconceptions.
- Feedback is provided.
- Students revise or improve their work.
This model is particularly appropriate for deep learning because learning is viewed as an ongoing process.
2. Authentic Assessment Model
Authentic assessment evaluates students through tasks that resemble real-world situations.
Students may be asked to:
- Solve community problems.
- Design a product.
- Conduct an investigation.
- Create a presentation.
- Develop a campaign.
- Analyze real data.
- Write a report.
- Conduct an experiment.
Authentic assessment can reveal whether students are able to transfer classroom knowledge into meaningful contexts.
3. Performance-Based Assessment
Performance-based assessment requires students to demonstrate a competency through action or performance.
For example:
- Conducting a scientific experiment
- Giving an oral presentation
- Performing a demonstration
- Creating a digital presentation
- Solving a practical problem
- Demonstrating a procedure
This model is useful when the learning objective involves practical skills, communication, collaboration, creativity, or problem-solving.
4. Project-Based Assessment
Project assessment evaluates learning through a structured project completed over a period of time.
A project may involve several stages:
- Identifying a problem.
- Planning the project.
- Collecting information.
- Developing a solution.
- Creating a product.
- Presenting the result.
- Reflecting on the process.
Project assessment can provide evidence of multiple competencies at once, including critical thinking, creativity, collaboration, communication, and independence.
5. Portfolio Assessment
A portfolio is a structured collection of students’ work that demonstrates learning progress over time.
A portfolio may include:
- Assignments
- Project results
- Essays
- Drawings
- Reports
- Reflection journals
- Revised work
- Certificates or evidence of activities
The important point is that a portfolio should not simply become a collection of files. Students should be encouraged to explain why particular works demonstrate their learning progress.
6. Peer Assessment Model
In peer assessment, students evaluate the work of their classmates based on agreed criteria.
This activity can develop:
- Critical thinking
- Communication
- Responsibility
- Collaboration
- Evaluation skills
To make peer assessment effective, teachers should provide clear criteria and teach students how to give constructive feedback.
7. Self-Assessment Model
Self-assessment encourages students to evaluate their own learning.
Students may assess:
- Their understanding of a concept
- Their participation
- Their strengths
- Their weaknesses
- Their learning strategies
- Their progress toward learning objectives
Self-assessment is particularly relevant to deep learning because it supports students’ ability to regulate their own learning.
Techniques for Assessing Learning Outcomes in Deep Learning
Different learning objectives require different assessment techniques. Therefore, teachers should select assessment techniques based on the competencies they want to measure.
1. Written Tests
Written tests remain useful for measuring knowledge and conceptual understanding.
They can include:
- Multiple-choice questions
- Short-answer questions
- Essay questions
- Problem-solving questions
- Case-based questions
To support deep learning, questions should not focus exclusively on memorization.
Instead of asking:
“Define photosynthesis.”
a deeper question could be:
“Explain how environmental changes can affect the process of photosynthesis and propose one strategy to reduce the impact.”
The second question requires students to understand, explain, analyze, and apply knowledge.
2. Oral Assessment
Oral assessment can be used to evaluate students’ ability to:
- Explain concepts.
- Express arguments.
- Answer questions.
- Communicate ideas.
- Defend a solution.
Teachers can use interviews, presentations, oral examinations, or classroom discussions.
3. Observation
Observation is useful for assessing behaviors and skills that may not be visible through written tests.
Teachers can observe:
- Participation
- Collaboration
- Communication
- Independence
- Problem-solving
- Practical skills
- Responsibility
Observation should ideally be supported by a checklist or rubric to make the assessment more systematic.
4. Project Assessment
Project assessment is particularly useful when students are expected to integrate knowledge and skills.
Teachers can assess both the process and the final product.
For example:
| Planning | Student develops a clear project plan |
| Investigation | Student collects relevant information |
| Collaboration | Student contributes effectively to the team |
| Creativity | Student develops an original solution |
| Product | Final product meets established criteria |
| Presentation | Student communicates findings clearly |
| Reflection | Student identifies strengths and areas for improvement |
5. Product Assessment
Product assessment focuses on something created by students.
Examples include:
- Posters
- Videos
- Models
- Reports
- Digital presentations
- Prototypes
- Infographics
- Learning media
The assessment should be based on clear criteria such as relevance, accuracy, creativity, quality, and usefulness.
6. Portfolio Assessment
Portfolio assessment can be used to show learning progression.
Teachers can compare students’ early work with later work to determine how their competencies develop.
This makes portfolio assessment especially useful for measuring progress rather than only final achievement.
7. Reflection
Reflection is an important technique in deep learning because students are encouraged to examine their learning experiences.
Reflection questions may include:
- What did I learn?
- What was the most important idea?
- What challenge did I face?
- How did I solve it?
- What evidence shows that I have improved?
- What do I still need to learn?
Reflection can be integrated into journals, learning logs, portfolios, or the end of a project.
Using Rubrics in Deep Learning Assessment
A rubric is one of the most useful tools for assessing complex learning outcomes.
A good rubric contains:
- Assessment criteria.
- Performance indicators.
- Performance levels.
- Clear descriptions of expected performance.
For example, a project rubric could include:
| Understanding | Demonstrates deep understanding | Demonstrates good understanding | Demonstrates partial understanding | Demonstrates limited understanding |
| Problem Solving | Develops an effective and well-supported solution | Develops an appropriate solution | Solution needs improvement | Solution is unclear |
| Creativity | Produces an original and innovative idea | Shows some originality | Limited originality | Mostly copies existing ideas |
| Communication | Communicates ideas clearly and convincingly | Communicates ideas clearly | Communication is sometimes unclear | Communication is difficult to follow |
| Reflection | Provides insightful reflection | Provides meaningful reflection | Reflection is limited | Reflection is minimal |
Rubrics help students understand what quality looks like before they begin a task. They also help teachers make assessment more transparent and consistent.
How to Design Assessment for Deep Learning
Effective assessment should begin with the learning objectives rather than with the assessment instrument.
A practical sequence is:
Step 1: Identify the Learning Objectives
Determine what students are expected to understand, demonstrate, create, or apply.
Step 2: Identify the Evidence
Ask:
What evidence would demonstrate that students have achieved the objective?
Step 3: Select the Assessment Technique
Choose a written test, observation, project, performance, portfolio, interview, self-assessment, or another appropriate technique.
Step 4: Develop Assessment Criteria
Define the characteristics of successful performance.
Step 5: Provide the Assessment Instrument
Prepare a rubric, checklist, question sheet, observation form, or other instrument.
Step 6: Conduct the Assessment
Collect evidence during or after the learning activity.
Step 7: Provide Feedback
Feedback should help students understand:
- What they have done well.
- What needs improvement.
- What they can do next.
Step 8: Encourage Reflection
Give students opportunities to evaluate their own learning.
Step 9: Use the Results
Assessment results should inform future teaching and learning decisions.
Examples of Deep Learning Assessment in the Classroom
Suppose students are learning about environmental pollution.
A conventional assessment might ask students to define pollution and identify several types.
A deep learning assessment could instead involve the following activity:
Learning Challenge:
Students investigate a pollution problem in their local environment and develop a practical proposal to address it.
Students could:
- Identify the problem.
- Collect information.
- Interview relevant people.
- Analyze possible causes.
- Develop a solution.
- Create a campaign or prototype.
- Present their findings.
- Receive peer and teacher feedback.
- Reflect on the learning process.
The teacher could assess:
- Knowledge and understanding
- Critical thinking
- Problem-solving
- Creativity
- Collaboration
- Communication
- Application of knowledge
- Reflection
This provides richer evidence than a single written examination.
Combining Assessment Techniques
There is no single assessment technique that can measure every dimension of student learning.
Therefore, teachers should combine several techniques.
For example:
Beginning of Learning
→ Diagnostic questions
→ Prior knowledge check
During Learning
→ Observation
→ Discussion
→ Formative quiz
→ Feedback
→ Peer assessment
End of Learning
→ Project
→ Presentation
→ Written assessment
→ Portfolio
→ Reflection
This combination provides a more comprehensive picture of learning.
The official Indonesian guidance for Pembelajaran Mendalam similarly provides examples including written tests, oral tests, performance assessment, project assessment, product assessment, observation, portfolios, peer assessment, and self-assessment.
Principles of Effective Assessment in Deep Learning
To make assessment meaningful, teachers should consider several principles.
1. Assessment Should Be Aligned
Assessment must be aligned with learning objectives and learning activities.
2. Assessment Should Be Continuous
Assessment should take place before, during, and after learning when appropriate.
3. Assessment Should Be Authentic
Whenever possible, students should be assessed through meaningful tasks and real-life contexts.
4. Assessment Should Encourage Reflection
Students should have opportunities to understand their strengths, challenges, and progress.
5. Feedback Should Be Actionable
Feedback should tell students not only what is wrong but also how they can improve.
6. Multiple Sources of Evidence Should Be Used
Teachers should avoid relying on a single test or instrument to make important judgments about learning.
7. Assessment Should Be Inclusive
Assessment should consider students’ diverse learning needs and provide appropriate accommodations when necessary.
8. Assessment Should Support Learning
Assessment should not become merely a mechanism for assigning grades. Its most valuable function is to provide information that improves learning.
Common Mistakes in Deep Learning Assessment
Teachers may encounter several challenges when implementing assessment.
Mistake 1: Focusing Only on Scores
A score alone cannot always explain why a student performs well or poorly.
Mistake 2: Using Only Written Tests
Written tests are useful, but they may not adequately measure creativity, collaboration, communication, practical skills, or reflection.
Mistake 3: Assessing Only the Final Product
The learning process is also important. Teachers should consider planning, investigation, collaboration, revision, and reflection.
Mistake 4: Giving Feedback Too Late
Feedback is most useful when students still have opportunities to use it to improve their work.
Mistake 5: Using Unclear Criteria
Students need to understand what is expected of them. Rubrics and clear criteria can make expectations more transparent.
Conclusion
Assessment in deep learning is more than measuring students’ final scores. It is a continuous process of collecting evidence, interpreting learning progress, providing feedback, encouraging reflection, and improving teaching and learning.
The three functions of Assessment as Learning, Assessment for Learning, and Assessment of Learning provide a strong framework for developing a comprehensive assessment system.
Teachers can combine various models and techniques, including formative assessment, authentic assessment, performance assessment, project assessment, portfolio assessment, peer assessment, self-assessment, written tests, oral assessments, observations, products, and reflection.
The most important principle is that assessment should answer a esensial question:
“How can the evidence we collect help students learn better?”
When assessment is designed with this purpose, it becomes more than a grading tool. It becomes an integral part of meaningful, reflective, and transformative learning.
Frequently Asked Questions (FAQ)
1. What is assessment in deep learning?
Assessment in deep learning is the process of collecting and using evidence of student learning to understand their progress, provide feedback, improve instruction, and measure achievement of learning objectives.
2. What are the three types of assessment in deep learning?
The three functions are Assessment as Learning, Assessment for Learning, and Assessment of Learning.
Assessment as Learning emphasizes student reflection and self-regulation. Assessment for Learning uses assessment information to improve ongoing learning. Assessment of Learning measures learning achievement, particularly at the end of a learning period.
3. What assessment techniques are suitable for deep learning?
Suitable techniques include written tests, oral assessments, observation, project assessment, performance assessment, product assessment, portfolios, peer assessment, self-assessment, and reflection.
4. Is a written test still appropriate for deep learning?
Yes. Written tests remain useful, particularly for assessing knowledge and conceptual understanding. However, they should ideally be combined with other assessment techniques when the learning objectives involve higher-order thinking, practical skills, creativity, collaboration, or communication.
5. Why is formative assessment important in deep learning?
Formative assessment provides information about student understanding while learning is still taking place. Teachers can use this information to provide feedback and adjust instruction according to students’ needs.
6. What is authentic assessment?
Authentic assessment evaluates students through meaningful tasks that resemble real-world situations. Students may solve problems, create products, conduct investigations, present findings, or develop solutions to practical challenges.
7. What is the role of a rubric in deep learning assessment?
A rubric provides clear assessment criteria and performance descriptions. It helps teachers assess complex tasks more consistently and helps students understand expectations and areas for improvement.
8. How can teachers assess critical thinking?
Critical thinking can be assessed through case studies, problem-solving tasks, projects, debates, investigations, open-ended questions, analysis activities, and reflection. The assessment should require students to explain reasoning rather than simply provide an answer.
9. How can students participate in assessment?
Students can participate through self-assessment, peer assessment, reflection journals, learning portfolios, goal setting, and progress checklists.
10. What is the main goal of assessment in deep learning?
The main goal is to obtain meaningful evidence of learning that can be used to understand student progress, improve teaching, encourage reflection, and support the development of competencies.
References
- Ministry of Primary and Secondary Education of the Republic of Indonesia. Panduan Pembelajaran dan Asesmen Pendidikan Anak Usia Dini, Jenjang Pendidikan Dasar, dan Jenjang Pendidikan Menengah – Edisi Revisi Tahun 2025. The guide explains the integration of learning and assessment and emphasizes formative assessment and competency development.
- Ministry of Primary and Secondary Education of the Republic of Indonesia. Pembelajaran Mendalam – Menuju Pendidikan Bermutu untuk Semua. The framework describes Assessment as Learning, Assessment for Learning, and Assessment of Learning and provides examples of assessment techniques.
- Ministry of Primary and Secondary Education of the Republic of Indonesia. Sistem Informasi Kurikulum Nasional – Pembelajaran Mendalam. The official framework describes deep learning as an approach emphasizing mindful, meaningful, and joyful learning.
- OECD. Using Formative Assessment and Feedback: Unlocking High-Quality Teaching. OECD explains the role of formative assessment, feedback, and the use of evidence to adapt instruction and support student progress.
- OECD. Student Assessment. OECD highlights the importance of assessment in understanding student progress and outcomes and balancing formative and summative functions.
- Ministry of Primary and Secondary Education of the Republic of Indonesia. Permendikdasmen No. 13 Tahun 2025. The regulation strengthens the direction toward implementing the deep learning approach in education and clarifies that it is an approach rather than a new curriculum.
