We finish reading a chapter and feel that its argument is clear. The examples are familiar, the terminology makes sense, and a second reading moves more quickly than the first. Yet a few days later, when someone asks us to explain the idea, we struggle to put it into words. The earlier sense of understanding was real as an experience, but it was an incomplete measure of what we could remember without the text in front of us (Bjork & Bjork, 2011).

Learning research helps explain this mismatch. Some conditions that make practice feel productive offer less support for later remembering than conditions that initially feel slower or less comfortable. Desirable difficulties are carefully chosen challenges that can improve long-term retention and the ability to apply knowledge. Their value depends on the mental activity they encourage, the learner’s existing knowledge, and what the learner will eventually need to do (Bjork & Bjork, 2011; Soderstrom & Bjork, 2015).

What Are Desirable Difficulties?

Desirable difficulties are conditions of learning that may make initial performance more demanding while improving long-term retention or transfer. Examples include spacing practice over time, recalling information from memory, interleaving related tasks, and varying practice conditions. These methods ask learners to do something beyond repeatedly encountering an immediately available answer (Bjork & Bjork, 2011).

The word desirable places an important limit on the idea. A challenge is useful when it engages processes that support learning and when the learner has sufficient knowledge or support to work with it. Confusing instructions or demands far beyond a learner’s preparation do not become beneficial simply because they are difficult. The purpose is to improve learning, not to maximize struggle (Bjork & Bjork, 2011).

Learning Versus Performance

Performance is what we can observe during practice: a correct answer, a clear explanation, or a successfully completed movement. Learning concerns changes in knowledge and skill that last beyond that moment. Researchers look for evidence of learning by checking what people remember later and what they can do in a different situation. Remembering a definition next week shows retention; recognizing how it applies in an unfamiliar example shows transfer (Soderstrom & Bjork, 2015).

An answer can come easily because it was just presented, because the same procedure has been repeated, or because the learning environment supplies useful cues. Those supports can improve present performance without ensuring that the answer will remain accessible when circumstances change. This does not make immediate success meaningless. It means that success during practice cannot, by itself, establish how durable learning has become (Bjork & Bjork, 2011; Soderstrom & Bjork, 2015).

In the Bjorks’ theory of memory, storage strength describes how firmly an idea is learned and connected to existing knowledge. Retrieval strength describes how easily it comes to mind at a particular moment. An idea we have just reread may be easy to recall even if we would struggle to remember it next week. The distinction helps explain why present confidence is an incomplete guide to lasting learning (Bjork & Bjork, 2011).

Judging learning requires us to interpret experiences that can be misleading. Familiar wording, quick answers, and a predictable sequence all offer reassuring signals. Retrieval and mixed practice can remove some of that reassurance, making progress feel less certain. Research on metacognition—the monitoring and regulation of our own thinking—shows that learners can mistake these differences in present performance for differences in lasting learning (Soderstrom & Bjork, 2015).

How Desirable Difficulties Support Learning

Spaced Practice: Returning to Learning Over Time

Spaced practice distributes encounters with material across time instead of concentrating them into an uninterrupted stretch. Returning to a concept after other activities have intervened changes the learning experience: the answer may no longer be immediately available, and the learner has another opportunity to reconstruct or relearn it. The Bjorks propose that reduced accessibility can create conditions for more substantial learning when the material is encountered again (Bjork & Bjork, 2011).

Cepeda and colleagues reviewed 317 experiments on remembering verbal information. They found that the most effective gap between study sessions depended partly on how much later people needed to remember the material. Longer gaps generally worked better when the final memory test was further in the future. Their review supports spacing practice over time, but it does not establish one ideal schedule for every subject and learner (Cepeda et al., 2006).

For someone learning a psychological theory, this could mean returning to its main ideas across several study sessions, rather than repeatedly reading the same explanation in one sitting. A later review can begin with an attempt to remember, followed by checking and clarification. Spacing describes when practice occurs; retrieval describes what the learner does during that practice. They can work together, but they are distinct features of learning (Bjork & Bjork, 2011; Cepeda et al., 2006).

Retrieval Practice: Remembering as a Learning Event

Retrieval practice involves bringing previously learned information to mind. Closing a book and explaining its central argument, answering a question before checking the solution, or building a concept map from memory can all create retrieval opportunities (Francis et al., 2020). The activity also reveals gaps, but its value goes beyond assessment: retrieving information can change how well it is remembered subsequently (Roediger & Karpicke, 2006).

In a well-known pair of experiments, Roediger and Karpicke asked students to study short prose passages and either restudy them or practice recalling them. Restudying favored performance on a test given five minutes later. At longer delays, including one week, prior retrieval produced better retention. Repeated reading also increased confidence, illustrating how a reassuring learning experience can diverge from later remembering (Roediger & Karpicke, 2006).

Classroom research supports the relevance of retrieval beyond laboratory tasks. A systematic review covering 50 experiments found benefits across several educational levels, subjects, and retrieval formats. Its reach still had limits: the evidence came predominantly from the United States and Western Europe, and the review excluded collaborative and online retrieval activities. The findings support retrieval as a useful educational practice without establishing equal benefits in every setting (Agarwal et al., 2021).

For practical use, a retrieval attempt should lead somewhere. After trying to explain an idea without the source, a learner can check what was accurate, identify what was missing, and revisit the explanation. Feedback makes this a process of correction rather than prolonged guessing. Retrieval does not need the pressure of a graded examination; accessible practice questions with answers available after an attempt can support its use (Carpenter, 2023).

Interleaving: Distinguishing Related Problems

Blocked practice groups similar tasks together: several problems using one procedure, followed by several using another. Interleaving mixes those problem types across a practice sequence. When a worksheet announces the procedure through its organization, a learner may successfully carry it out without having to decide when it applies. A mixed sequence can add that decision, requiring attention to the features that distinguish one problem from another (Bjork & Bjork, 2011).

However, interleaving is not uniformly superior. Brunmair and Richter’s meta-analysis found an overall advantage, with stronger results for visual category learning, such as distinguishing styles of paintings. Results were less consistent across other materials, and some word-learning tasks favored blocking. Their findings support attention to the material and the intended comparison, rather than a rule that all topics should always be mixed (Brunmair & Richter, 2019).

Varied Practice Across Examples and Conditions

Varied practice changes aspects of the examples, demands, or circumstances in which a skill is used. A learner might encounter a principle in several different situations rather than repeatedly seeing the same example. Research reviewed by Soderstrom and Bjork suggests that variation can sometimes improve later performance under changed conditions, even when it introduces more errors during acquisition. One explanation is that learning becomes connected to a broader range of relevant cues (Soderstrom & Bjork, 2015).

Interleaving and variation can overlap, but they answer different questions. Interleaving concerns the order of related tasks; variation concerns what changes across practice. Neither requires constant interruption or unrelated distraction. A useful application is to vary examples in ways that preserve the underlying principle while helping the learner recognize it in new circumstances. Success with those examples should still be checked rather than assumed (Bjork & Bjork, 2011; Soderstrom & Bjork, 2015).

Why Effective Learning Strategies Can Be Hard to Sustain

The appeal of comfortable study routines has a connection with the law of least effort. Yet understanding that retrieval is effective does not guarantee that students will choose it. Carpenter’s review found that information about retrieval benefits, and even opportunities to experience retrieval, were not reliably sufficient to change study choices. Perceived difficulty and errors could discourage its use (Carpenter, 2023).

Making retrieval more approachable can encourage students to use it. In some studies, hints, opportunities to succeed, and feedback showing learning benefits increased students’ willingness to choose retrieval practice. These findings bring motivation into the discussion: even an effective method has limited practical value if learners repeatedly avoid it. A useful challenge must be one they can engage with and return to (Carpenter, 2023).

The Start and Stick to Desirable Difficulties framework extends this discussion by examining how perceived effort and perceived learning shape continued engagement. A learner may interpret a demanding retrieval attempt as evidence that the method is ineffective, then abandon it before experiencing its benefits. The framework proposes supporting both accurate expectations and practical experience with strategies, including feedback on their outcomes. It is a research framework rather than a fully validated intervention (de Bruin et al., 2023). It directs attention to an important question: what helps people sustain a useful learning practice?

When Difficulty Stops Helping

Prior knowledge changes what a learner can do with a challenge. A person who understands the basic concepts may benefit from having to retrieve and distinguish them. Someone who has not yet grasped those concepts may need an explanation or a worked example before attempting the same task (Sweller, 1988; Chen et al., 2018). The Bjorks explicitly recognize this boundary: without sufficient background knowledge or skill, a desirable difficulty can become an undesirable one (Bjork & Bjork, 2011).

Working hard to solve a problem does not always help us learn how to solve the next one. Sweller’s research on cognitive load helps explain why: the search for an answer can use mental resources that are also needed for learning (Sweller, 1988). Chen and colleagues argue that extra difficulty may hinder learning when a task already requires us to hold and connect several unfamiliar ideas in working memory. They call this element interactivity. Its demands depend on both the material and what the learner already understands. In such cases, a clear explanation or a step-by-step worked example may be more helpful than another attempt to recall or work out an answer without enough support (Chen et al., 2018).

The form of retrieval practice also matters. In a study of 53 introductory psychology students, Francis and colleagues found that students scored better on exam material they had practiced by building concept maps from memory. This pattern appeared among students who began with both more and less knowledge of psychology. Multiple-choice practice showed an advantage over material without organized in-class retrieval practice only among students with more prior knowledge. Students received instructions for making the maps and feedback afterward. However, different topics were assigned to different practice methods, so differences in the material could partly explain the results. The study therefore cannot establish that concept mapping is generally superior (Francis et al., 2020).

These boundaries also matter when interpreting errors. An unsuccessful attempt can identify where more support is needed, but failure alone does not demonstrate that learning has occurred. The relevant question is whether subsequent explanation, correction, and practice improve what the person can remember or do. Neither ease nor frustration should become the sole measure of educational value (Bjork & Bjork, 2011; Soderstrom & Bjork, 2015).

A practical way to judge a learning challenge is to consider what it asks the learner to do, what support is available, and what the learner can remember or apply later. The comparison below draws these questions together as a Psychology Fanatic synthesis of the research. Its examples illustrate how support can be adjusted to the demands of a task (Bjork & Bjork, 2011; Chen et al., 2018; de Bruin et al., 2023).

Retention and Transfer: Remembering and Applying Ideas

Remembering and understanding are connected, but a familiar answer does not demonstrate every kind of understanding. A learner might accurately state that correlation does not establish causation, yet accept a causal claim from a study that only reports an association. Explaining why the claim exceeds the evidence requires the learner to use the principle, not simply repeat its wording. This example illustrates why retention and transfer need to be examined separately (Soderstrom & Bjork, 2015).

Pan and Rickard’s review combined results from many experiments to ask whether practice tests help people use what they learn in new situations. Compared with studying the material again, practice tests often helped, but the benefit depended on the later task. Better results were associated with recalling answers successfully during practice, explaining or expanding on those answers, and practicing responses that were relevant to the later questions. These findings suggest that remembering can support flexible use of knowledge, while reminding us that practicing one answer does not prepare us for every new problem (Pan & Rickard, 2018).

One implication is to align practice with the way knowledge will eventually be used (Pan & Rickard, 2018). Applied to evaluating research, this suggests giving learners opportunities to explain evidence, compare possible interpretations, and assess unfamiliar examples. Later assessment would examine those abilities as well as recall. Merely changing the wording of a familiar question tells us less about whether a learner can recognize when a principle applies.

Putting the Principles into Practice

Consider a learner returning to the correlation-and-causation example. An instructor first explains the distinction through a clear case, including how a third factor could account for an association. The learner then closes the explanation and reconstructs the reasoning. If the distinction is unclear, another worked example or a hint can provide support before the learner tries again. This sequence illustrates how guidance and retrieval can complement one another (Bjork & Bjork, 2011; Chen et al., 2018).

On a later day, the learner revisits the idea through new descriptions of studies. Some report associations; others describe randomized experiments. Comparing them requires attention to how evidence was produced and what conclusions it supports. Spacing separates the encounters, retrieval reconstructs prior learning, and varied examples invite application. Mixing study types also creates an opportunity for comparison, although the success of that arrangement should be evaluated for this particular material rather than inferred from interleaving research alone (Bjork & Bjork, 2011; Brunmair & Richter, 2019).

The instructor can then examine whether the learner can explain a conclusion without the original example and identify its limits in another case. The learner can also compare present understanding with an earlier attempt, making progress more visible. If answers remain confused, the next step may be clearer instruction or simpler examples. If performance improves only on familiar questions, practice may need to address application more directly. Difficulty remains useful only insofar as it helps the learner move toward the intended outcome (Soderstrom & Bjork, 2015; Carpenter, 2023).

A Few Words by Psychology Fanatic

The moment we close a book can reveal something that an easy reading concealed. We may discover an idea we can explain, a connection we only partly understand, or a question that deserves another look. Such discoveries need not be treated as verdicts on our ability. They can help us decide what the next encounter with the material should offer.

We do not need to distrust every experience of ease or admire every struggle. We can allow learning to include clear explanations, patient repetition, and occasions when remembering takes work. Over time, the more meaningful reassurance comes when an idea remains available after the page is gone—and when we can recognize where it belongs in a situation we have not seen before.

Associated Concepts

• Metacognition: Monitoring what we know and using that information to adjust how we study.
• Law of Least Effort: A related perspective on why immediately comfortable ways of studying can be appealing.
• Cognitive Load Theory: How task demands and prior knowledge influence the mental effort available for learning.
• Working Memory: The limited capacity to hold and work with information, helping explain why added difficulty can sometimes overwhelm a learner.
• Self-Regulated Learning: How learners plan, monitor, and adjust study strategies using feedback.
• Feedback Loops: How information about performance guides adjustments to later practice.

References

Agarwal, Pooja K.; Nunes, Ludmila D.; Blunt, Janell R. (2021). Retrieval Practice Consistently Benefits Student Learning: A Systematic Review of Applied Research in Schools and Classrooms. Educational Psychology Review, 33, 1409–1453. DOI: 10.1007/s10648-021-09595-9.
(Return to Main Text)

Bjork, Elizabeth Ligon; Bjork, Robert A. (2011). Making Things Hard on Yourself, but in a Good Way: Creating Desirable Difficulties to Enhance Learning. In M. A. Gernsbacher, R. W. Pew, L. M. Hough, & J. R. Pomerantz (Eds.), Psychology and the Real World: Essays Illustrating Fundamental Contributions to Society (pp. 56–64). Worth Publishers. Website: UCLA author-hosted chapter.
(Return to Main Text)

Brunmair, Matthias; Richter, Tobias. (2019). Similarity Matters: A Meta-Analysis of Interleaved Learning and Its Moderators. Psychological Bulletin, 145(11), 1029–1052. DOI: 10.1037/bul0000209.
(Return to Main Text)

Carpenter, Shana K. (2023). Encouraging Students to Use Retrieval Practice: A Review of Emerging Research from Five Types of Interventions. Educational Psychology Review, 35, Article 96. DOI: 10.1007/s10648-023-09811-8.
(Return to Main Text)

Cepeda, Nicholas J.; Pashler, Harold; Vul, Edward; Wixted, John T.; Rohrer, Doug. (2006). Distributed Practice in Verbal Recall Tasks: A Review and Quantitative Synthesis. Psychological Bulletin, 132(3), 354–380. DOI: 10.1037/0033-2909.132.3.354.
(Return to Main Text)

Chen, Ouhao; Castro-Alonso, Juan C.; Paas, Fred; Sweller, John. (2018). Undesirable Difficulty Effects in the Learning of High-Element Interactivity Materials. Frontiers in Psychology, 9, Article 1483. DOI: 10.3389/fpsyg.2018.01483.
(Return to Main Text)

de Bruin, Anique B. H.; Biwer, Felicitas; Hui, Luotong; Onan, Erdem; David, Louise; Wiradhany, Wisnu. (2023). Worth the Effort: The Start and Stick to Desirable Difficulties (S2D2) Framework. Educational Psychology Review, 35, Article 41. DOI: 10.1007/s10648-023-09766-w.
(Return to Main Text)

Francis, Andrea P.; Wieth, Mareike B.; Zabel, Kevin L.; Carr, Thomas H. (2020). A Classroom Study on the Role of Prior Knowledge and Retrieval Tool in the Testing Effect. Psychology Learning & Teaching, 19(3), 258–274. DOI: 10.1177/1475725720924872.
(Return to Main Text)

Pan, Steven C.; Rickard, Timothy C. (2018). Transfer of Test-Enhanced Learning: Meta-Analytic Review and Synthesis. Psychological Bulletin, 144(7), 710–756. DOI: 10.1037/bul0000151.
(Return to Main Text)

Roediger, Henry L., III; Karpicke, Jeffrey D. (2006). Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention. Psychological Science, 17(3), 249–255. DOI: 10.1111/j.1467-9280.2006.01693.x.
(Return to Main Text)

Soderstrom, Nicholas C.; Bjork, Robert A. (2015). Learning Versus Performance: An Integrative Review. Perspectives on Psychological Science, 10(2), 176–199. DOI: 10.1177/1745691615569000.
(Return to Main Text)

Sweller, John. (1988). Cognitive Load During Problem Solving: Effects on Learning. Cognitive Science, 12(2), 257–285. DOI: 10.1207/s15516709cog1202_4.
(Return to Main Text)