Rethinking Testing: Better Ways to Use Assessment to Improve Learning (Part 7/7)
Phew! You made it! This is the seventh story in a series of posts about how learning science insights can improve course design. The other stories are linked within and included at the end of this post.
Teaching to the test or testing to teach?
Ah, assessment. Tests, quizzes, essays, projects, presentations, reports. They cause all-nighters, headaches, tears. They dominate teacher time and parents’ understanding of success.
And they can absolutely be improved through learning science.
We’ll talk about formative assessment (assessment for learning) and summative assessment (assessment of learning). We’ll talk about aligned assessment and authentic assessment. But first, let’s talk about testing.
What do we mean by testing? Many people think of a closed-book test, assessing students’ knowledge of facts, concepts and ideas, resulting in a grade. Learning science, on the other hand, rethinks testing.
Instead of being a final assessment activity, psychology shows that testing is among the most effective learning activities. Here’s how:
Testing Strengthens Memory
Most people think repetition improves memory. Read your notes several times, solve several math worksheets, review your vocabulary and spelling words. What this typically does, however, is produce what’s called a ‘fluency illusion’ [1], where we feel like we know material because we’re so familiar with it, but when we come to an assessment, we can’t recall the details.
Instead, we now know that the process of trying to recall information improves memory. We shouldn’t view learning as the process of putting things in our memory, but rather as the process of retrieving things from our memory.
How does this work? When students study information several times, they do well on immediate tests, but poorly on tests later on. This suggests they haven’t retained the material. By contrast, when students study information just a couple of times, but test themselves on what they learned (for example, through self-quizzing or trying to write down a text from memory) their long-term retention is significantly better.
Testing works best for long-term retention when it’s done repeatedly and at ever increasing intervals [2]. So quizzing students five times on the same day is less effective than quizzing students the day after learning, then two days later, then a week later, 2 weeks later, and a month later, for example.
Interestingly, the format of the test used for retrieval practice can be very different from the actual final test used to assess learning. The act of retrieving the information from memory seems to strengthen the memory which is then more easily accessible in the future, no matter the assessment context [3].
This has been tested with a wide range of subject matter: foreign vocabulary words, word lists, picture lists, reading exercises, physical processes (such as the steps to take in resuscitation), math facts, factual information from school curricula and discipline-specific terminology.
Testing, therefore, should be a key component of our learning science-driven curriculum planning.
Tying It Together
It’s time for us to move testing out of the realm of assessment and embrace it as a key learning strategy.
The more we see how well we know something by actually forcing ourselves to generate a response, the better our ability to retrieve that information later. When we need it, we’ll know it.
Testing and Assessment
So how does this fit with our assessment planning? The first step, as in many curriculum models, is to define the assessment aligned with the learning outcomes [4]. Why? This ensures we actually measure what we set out to teach. If we claim we have conceptual learning outcomes, but create factual recall assessments, there’s a mismatch that students will quickly pick up on.
Within this step, we consider the type of assessment. Of course we have standard tests and exams, but as teachers we know there are many other ways to evaluate whether students have learned the skills and concepts.
Authentic assessments — written work that is published to a broader audience, media presentations that are shared beyond the classroom, projects that result in organizational change — are powerful motivators. Among the diverse theories of motivation, three consistent elements that stand out are people’s desire to control their situation, demonstrate their skill, and people’s desire to be part of a group [5]. Authentic assessments motivate students by drawing on each of these elements.
But no matter what the format of our assessment, we can, and should, take advantage of the testing effect to improve learning towards those learning goals. This is where formative assessment comes in.
Assessment for Learning
Typically, we consider formative assessments to be a rich teaching tool, providing the teacher with feedback about misconceptions and confusion and helping the teacher review key concepts and ideas [6].
Given the power of the testing effect, we now know that formative assessment is key for students’ actual learning. Viewed through the lens of the testing effect, formative assessment is actually causing learning.
This dual benefit then — that formative assessment provides valuable feedback to teachers to redirect their teaching AND improves long-term retention of information — makes formative assessment an even more important element of our teaching practice.
Mapping Assessment
One last thing about assessment. Often, we think about assessment just in terms of one unit of study. The fractions test. The French Revolution report. The To Kill a Mockingbird essay. Learning science suggests we take a longer view and map assessments throughout a course/year to follow the red thread throughout.
There are two key takeaways from learning science that can frame how we look at assessment:
- Feedback research which emphasizes the importance of providing feedback that lets students know how they are currently performing and what they need to do in order to progress [7].
- Psychology research which emphasizes the importance of interleaving concepts so that assessments target new learning as well as previously learned and assessed material [3].
There’s a whole post about feedback, so I won’t repeat much here, but what’s relevant to this argument is that your assessments should be planned so that students can use your feedback to gradually build up skills.
Does this mean that every assessment targets the same skills? Absolutely not. But some of the concepts or skills or processes should be clearly linked to the prior work and feedback they’ve received.
In what seems like an apparent contradiction, psychology research strongly recommends interleaving for improving learning, long-term retention, and transfer. Interleaving basically means mixing things up.
Here’s an example: In a math class students take a unit test assessing specific skills and then sit for a final exam that mixes up the skills taught throughout the semester. Typical final exams don’t have Section A: adding fractions, Section B: multiplying fractions, Section C: converting from fractions to decimals. Instead, they typically have the three types of questions mixed up, with the idea that you’re assessing not only whether students know the skill, but also know when to apply the skill.
The problem is, the typical unit test doesn’t prepare them for that. Typical homework assignments and unit tests address skills discretely, so students practice solving problems with the quadratic equation, but don’t ever think about what types of problems require them to apply the quadratic equation. When the worksheet is titled “Quadratic Equations” or the unit test focuses solely on the quadratic formula, students just apply the process they learned.
Interleaving, or mixing up practice and tasks required on formative and summative assessments, improves learning precisely because it requires students to not only learn the process, but also learn how to discriminate between processes and determine which strategy/equation/concept to apply to which question.
Tying It Together
The learning science take-away? Plan your assessments so that there’s a consistent element which students can use prior feedback to build up their skills and abilities on AND make sure there’s a diversity of types of tasks or concepts assessed so that students are practicing not only knowing the concept, but also knowing when the concept is relevant.
The Take-Away
So, learning science-driven assessment in a nutshell:
- Plan the assessment in advance to be aligned with the learning outcomes so you actually measure what you intend to teach.
- Draw on theories of motivation to plan assessments that provide students some element of control (or choice), that allow them to demonstrate their developing skills, and that encourage them to feel part of a larger group. In other words, switch things up so sometimes students do a writing assignment, sometimes it’s a straightforward test, sometimes it’s a project, sometimes it’s shared with the broader community, sometimes it’s collaborative, sometimes it’s creative.
- The testing effect just doubled the power of formative assessment. It’s a valuable tool for teachers to intervene when students struggle (or excel) and it’s a priceless tool for students in improving long-term learning.
- Plan your assessments so that students can apply your feedback to the next task.
- Create your assessments so that they challenge students to demonstrate understanding and also determine when to apply various concepts.
Related Posts
How Insights from Learning Science Can Transform Your Teaching (Part 1/7)
Three Insights From Learning Science to Structure Your Lessons Better (Part 2/7)
Three Powerful Lessons from Psychology To Change How You Plan Lesson Content (Part 3/7)
What Learning Science Says About How to Teach (Part 4/7)
How You Can Use a Top-Ten Instructional Strategy to Boost Learning (Part 5/7)
The Surprising Ways Thinking About Learning Can Impact Learning (Part 6/7)
Teaching in a Pandemic: How Learning Science can Help (Part 8/7)
Prior Knowledge: Why It Matters and What We Can Do
References
- Karpicke, J. D., Butler, A. C., & Roediger III, H. L. (2009). Metacognitive strategies in student learning: do students practise retrieval when they study on their own?. Memory, 17(4), 471–479.
- Pyc, M. A., Agarwal, P. K. & Roedinger III, H. L. (2014). Test-enhanced learning. In V. A. Benassi, C. E. Overson & C. M. Hakala (Eds), Applying science of learning in education: Infusing psychological science into the curriculum (pp. 59–70). Retrieved from the Society for the Teaching of Psychology website: http://teachpsych.org/ebooks/asle2014/index.php
- Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students’ learning with effective learning techniques: Promising directions from cognitive and educational psychology. Psychological Science in the Public Interest, 14(1), 4–58.
- Biggs, J. (1996). Enhancing teaching through constructive alignment. Higher education, 32(3), 347–364.
- Anderman, L. H., & Leake, V. S. (2005). The ABCs of Motivation: An Alternative Framework for Teaching Preservice Teachers about Motivation. Clearing House: A Journal of Educational Strategies, Issues and Ideas, 78(5), 192.
- Wiliam, D. (2011). Embedded formative assessment. Solution Tree Press.
- Hattie, J., & Timperley, H. (2007). The power of feedback. Review of educational research, 77(1), 81–112.
If you enjoyed this piece, hold the 👏 icon below so others can find this story. Please share your thoughts in the comments.
For more pieces from Learn Better, follow us on our page.
Stephanie Hepner has taught middle and high school special education/learning support and English in New York, Brussels, and Stockholm. She currently works in education in Singapore. An international educator committed to equity in education, she is passionate about learning science as it promises to improve learning for all students.
