The Processing Difficulty Effect

At a Glance

1. Quick Summary

The processing difficulty effect is the finding that information we work harder to understand is often remembered better than information that slides in easily. In 1985 Edward O'Brien and Jerome Myers showed that readers who had to make an extra inference to connect a passage recalled more of it, and Robert Bjork's idea of "desirable difficulties" gave the pattern a name and a theory. The trouble is the word "difficulty," because there are two very different kinds. Difficulty that makes you think about meaning (producing an answer, explaining, connecting ideas) improves memory in study after study. Difficulty that only makes the letters harder to see, such as an ugly font, briefly looked like a cheap learning trick after a few small experiments between 2007 and 2011, and then failed in large replications, a meta-analysis and several tests of a font designed for the purpose. The useful lesson is to make learning effortful in the right place, and to stop trusting how easy something feels as a sign that you have learned it.

Simply About the Bias

The processing difficulty effect means that when your brain has to work a bit to figure something out, you usually remember it better. If the answer is handed to you, it goes in easily and slips out easily. If you had to dig for it, it tends to stay.

Think about learning the rules of a new board game. Your cousin reads the rulebook out loud while you half listen, and an hour later you can't remember how scoring works. The next week you have to teach the game to your friends, so you sit down and work out each rule yourself, trying little examples. You remember that version for months. The second time felt harder, and that's exactly why it stuck. There's a catch, though. Making the rulebook blurry or printing it in a weird font wouldn't have helped. The hard part has to be the thinking, not the squinting.

Here's a trick. After you read a page of your notes, close them and try to say the main points out loud, or write down three questions and answer them without looking. If it feels a bit hard, good. That's the kind of difficulty that helps you remember.

2. The Science Behind It

2.1. Discovery and History

The Cognitive Bias Codex lists the processing difficulty effect as the tendency for information that takes longer to read, and that we think about more, to be remembered more easily. Researchers use other names (desirable difficulties, depth of processing, the generation effect, disfluency), and these strands differ sharply in how well they hold up.

The deep-processing line came first. In 1972 Fergus Craik and Robert Lockhart argued that attending to what a word means leaves a stronger memory trace than attending to how it looks or sounds. Norman Slamecka and Peter Graf (1978) then showed that people remember words they produce themselves better than words they simply read. A 2007 meta-analysis by Bertsch, Pesta, Wiscott and McDaniel put the average advantage at roughly 0.4 of a standard deviation across 86 studies.

Edward O'Brien and Jerome Myers carried the idea to connected text in 1985. Readers who hit a point in a passage that was harder to integrate slowed down, and later recalled more of those passages. The extra time, they argued, went into searching back through the text to make the inference, which wove more of the passage together in memory.

Robert Bjork's 1994 chapter on training pulled these results into a framework. Spacing practice, mixing problem types and testing yourself feel slow and error-prone, yet often produce better long-term retention than smooth study. He called such conditions "desirable difficulties," and the qualifier mattered: a difficulty helps only if the learner can overcome it and doing so engages the processes that build memory.

The second line, perceptual disfluency, came from social psychology. Adam Alter, Daniel Oppenheimer, Nicholas Epley and R. N. Eyre (2007) reported that printing reasoning problems in a degraded font made people answer more analytically. Connor Diemand-Yauman, Oppenheimer and Erikka Vaughan (2011) reported that hard-to-read fonts improved retention in a lab and in high-school classrooms. Both were small studies, and both spread fast.

Then the replications arrived. Andrew Meyer, Shane Frederick and colleagues (2015) pooled the original font-and-reasoning experiment with 16 attempts to repeat it, more than 7,000 participants in all, and found no effect. In 2016, Ralf Rummer, Judith Schweppe and Annett Schwede, and separately Alexander Eitel and Tim Kühl, reported null effects on learning. He-Ping Xie, Zongkui Zhou and Qing-Qi Liu's 2018 meta-analysis concluded that perceptual disfluency had no reliable benefit for recall or transfer from text. Sans Forgetica, a font built at RMIT University to be "desirably difficult," was tested independently in 2020 and showed no memory benefit. A 2026 study by Muhmenthaler and Meier at the University of Bern found a hard font helped recall only in an easy two-choice task, and concluded the benefit "may be difficult to reliably elicit."

Key milestones:

• 1917: Arthur Gates finds that schoolchildren who spend much of their study time reciting from memory recall more than those who mostly reread
• 1972: Craik and Lockhart propose the levels-of-processing framework
• 1978: Slamecka and Graf publish the generation effect
• 1985: O'Brien and Myers show that text requiring more inference is recalled better
• 1994: Bjork names "desirable difficulties"
• 2007: Alter and colleagues report that a degraded font improves reasoning
• 2011: Diemand-Yauman, Oppenheimer and Vaughan report better retention with hard-to-read fonts
• 2015: Meyer and colleagues pool 17 studies and find no font benefit on reasoning
• 2018: Xie, Zhou and Liu's meta-analysis finds no reliable learning benefit of perceptual disfluency; Sans Forgetica is launched
• 2020: Two independent teams find no memory benefit from Sans Forgetica

2.2. Key Researchers

2.3. Landmark Studies

"The Generation Effect: Delineation of a Phenomenon" (Slamecka & Graf, 1978)

In the read condition participants saw a full pair such as "rapid: fast." In the generate condition they saw "rapid: f___" and a rule telling them to produce a synonym. Across five experiments and several rules, generated words were remembered better, a result that has held up for decades.

"When Comprehension Difficulty Improves Memory for Text" (O'Brien & Myers, 1985)

People read short passages in which a target concept was either easy or harder to connect to the preceding context. Reading slowed where integration was harder, and memory for those passages was better. The print never changed; the difficulty sat in the meaning.

"Overcoming Intuition" (Alter, Oppenheimer, Epley & Eyre, 2007)

Forty Princeton students answered the three-item Cognitive Reflection Test, which includes the bat-and-ball problem, printed either clearly or in small, grey, hard-to-read type. In the clear condition 90% got at least one item wrong; in the degraded condition, 35% did. The authors proposed that difficulty works as an alarm that switches on analytic thinking, and Daniel Kahneman retold the result in Thinking, Fast and Slow (2011).

The Pooled Replication (Meyer et al., 2015)

Meyer and colleagues combined the original data with 16 replications run by different teams, more than 7,000 participants in total. The hard font had no effect on solution rates. They found no moderator, whether cognitive ability, format or setting, under which it helped. The font slowed people slightly and did not make them reason better.

"Fortune Favors the Bold (and the Italicized)" (Diemand-Yauman, Oppenheimer & Vaughan, 2011) and Its Replications

In the lab study, 28 participants learned facts about fictional alien species. Those who read them in harder, lower-contrast fonts recalled about 86.5% against 72.8% for the clear-font group. A high-school follow-up with worksheets in fonts such as Haettenschweiler reported somewhat better test scores. Rummer, Schweppe and Schwede (2016) found no benefit in a close replication, Eitel and Kühl (2016) found none whether or not learners expected a test, and the 2018 meta-analysis found no reliable effect.

The Sans Forgetica Tests (Taylor, Sanson, Burnell, Wade & Garry, 2020; Geller, Davis & Peterson, 2020)

Sans Forgetica slants backwards and has gaps cut out of each letter. Taylor and colleagues ran four experiments: people rated it harder to read than Arial, but memory was the same and sometimes worse. Geller, Davis and Peterson's first preregistered experiment is the cleanest demonstration of the whole topic. Targets were studied either in Sans Forgetica or with letters missing (G_RL), which forces generation. Generation produced a robust memory benefit. The font produced none. Huff, Maxwell and Mitchell (2022) found nothing in four more experiments.

2.4. Neurological Basis

There is no established neural signature specific to "processing difficulty," and claims that a hard font activates particular brain regions to improve memory lack solid evidence. Neuroscience does support the deep-processing half of the story. Imaging studies of encoding have found that judging a word's meaning, compared with judging its letters or sound, engages the left inferior prefrontal cortex more strongly (Kapur and colleagues, 1994). Anthony Wagner and colleagues (1998, Science) showed that activity in left prefrontal and medial temporal regions while a word is studied predicts whether it will later be remembered.

Perceptual difficulty mostly adds work at the level of visual word recognition. It slows reading, but a reader can spend the extra milliseconds decoding letters without thinking any harder about content.

3. Evolutionary Origins

No one has tested an evolutionary account of this effect directly, so what follows is reasoning rather than evidence.

Effort as a relevance signal: Information you had to work out yourself, such as which route avoided a predator, was usually worth keeping, so weighting effortful episodes more heavily would keep the right things most of the time.

Generated knowledge is connected knowledge: Inferring or generating something pulls existing knowledge into play. The new item ends up attached to many old ones, each a route back to it.

Fluency as a familiarity cue: Ease of processing has long been a decent sign that something is familiar. That cue misleads in a world of slides and textbooks, where fluent presentation is cheap and says nothing about whether you have learned the content.

Why the font trick fails: Ancestral difficulty came from the problem, not the display. A smudged mark was not more important for being smudged, so there was no reason for the brain to reward visual awkwardness.

4. How This Bias Manifests

4.1. In Everyday Life

• Remembering the way to a friend's house after driving there yourself, but not after riding as a passenger
• Rereading a chapter until it feels easy and mistaking that ease for knowing it
• Remembering a recipe you improvised better than one you followed step by step
• Buying a "study hack" font or app on the promise that harder-to-read text will improve memory

4.2. In the Workplace

• New hires who must do a task and explain each step remember more than those who shadow a colleague
• Documents printed in awkward fonts "so people read carefully," which mostly slows and annoys readers
• Note-taking debates: Pam Mueller and Daniel Oppenheimer (2014) reported that longhand notes beat laptop notes because handwriting forces summarising, but a larger replication by Kayla Morehead, John Dunlosky and Katherine Rawson (2019) did not find a reliable advantage

4.3. In Business and Marketing

Marketers mostly need the opposite of disfluency. Hyunjin Song and Norbert Schwarz (2008) found that exercise instructions printed in a hard-to-read font were judged to take longer and need more skill, which made people less willing to try the routine.

• Ads that ask the audience to complete a phrase or solve a small puzzle, which is closer to generation than to disfluency
• Learning products sold as "brain training" because they feel hard, without evidence the difficulty transfers

4.4. In Politics and Media

Fluency helps falsehoods: easy-to-read, repeated statements are judged more likely to be true. Song and Schwarz (2008) reported that a harder font reduced the "Moses illusion" (missing a planted error in a question), but the large replications argue against using fonts to fight misinformation.

Manifestations include:

• Voters who skim headlines and confuse familiarity with understanding
• Fact-checks that are read passively and forgotten while the fluent original claim lingers

4.5. In Healthcare

Patient instructions should be easy to read, since disfluent text discourages people from following them. Remembering them is helped by effortful processing of meaning.

• The "teach-back" method, in which patients explain their care plan in their own words
• Residents who learn with practice questions retain more than those who reread; Douglas Larsen and colleagues reported this in randomised trials
• Simulation training, where trainees must decide and act, remembered better than lectures on the same material

4.6. In Finance and Investing

• Investors who build their own model of a company remember its numbers and risks better than those who read an analyst summary
• Easy-to-pronounce company names: Adam Alter and Daniel Oppenheimer (2006) found that stocks with more pronounceable names performed better shortly after listing, a sign of how fluency shapes market judgment

5. Real-World Case Studies

Case Study 1: Sans Forgetica

• Context: In 2018, researchers and designers at RMIT University in Melbourne released Sans Forgetica, a free font with back-slanted letters and gaps cut into each character, presented as a study aid based on desirable difficulty.
• What happened: The launch drew international press and wide downloads. The supporting evidence was unpublished in-house testing. In 2020, Taylor and colleagues found that people rated the font harder to read, but it did not improve memory and sometimes hurt it. Geller, Davis and Peterson set it against a generation task: generation helped, the font did nothing. Huff, Maxwell and Mitchell (2022) found no effect in four more experiments.
• The bias at work: The project took a real effect (effortful processing of meaning improves memory) and applied it to the wrong kind of effort. Harder letters add work at the level of seeing, not thinking.
• Consequences: Students studied in a font that offered no benefit, and the episode became a teaching example of a theory stretched past its evidence.
• Lessons learned: Ask where an intervention puts the effort. Wait for independent, preregistered tests before adopting a learning product.

Case Study 2: The Hard Font and the Bat-and-Ball Problem

• Context: The Cognitive Reflection Test asks questions with tempting wrong answers. A bat and ball cost $1.10 and the bat costs $1.00 more: how much is the ball? (The intuitive answer is 10 cents; the right one is 5.)
• What happened: Alter and colleagues (2007) reported that printing these questions in a small, grey font cut the share of people making at least one error from 90% to 35%, in a sample of 40. In 2015 Meyer and colleagues pooled it with 16 replications, more than 7,000 people, and found no effect.
• The bias at work: A small sample produced an implausibly large effect that fitted a neat story about effort and attention.
• Consequences: For years, talks and training sessions taught "use hard fonts to make people think" as established science.
• Lessons learned: A big effect from a small sample is a reason for caution. Corrections tend to travel less far than the original claim.

Historical Example: Arthur Gates and Recitation (1917)

In 1917 Arthur Gates at Columbia University published "Recitation as a factor in memorizing." Schoolchildren studied short biographies and nonsense syllables, splitting their time differently between reading and reciting with the text out of sight. Those who spent a large share reciting remembered more. Reciting is harder: you stumble, fail to recall, check and try again.

This is an early, clean case of the effect in its reliable form, now studied as retrieval practice or the testing effect. Henry Roediger and Jeffrey Karpicke revived it in 2006, showing that students who practised recalling a passage remembered more a week later than students who reread it, even though the rereaders felt more confident.

6. The Cost of This Bias

6.1. Personal Costs

• Hours of rereading and highlighting with little to show at exam time
• Avoiding effortful study methods because struggle feels like evidence of not learning
• Overconfidence before tests after material went in smoothly

6.2. Professional Costs

• Training budgets spent on polished content that is consumed passively and forgotten
• Staff who sat through safety training but cannot recall it when it matters
• Trainers adopting hard fonts on the strength of one viral study

6.3. Societal Costs

• Education that favours coverage and clear lecturing over retrieval and problem-solving, producing knowledge that fades quickly
• Fluent misinformation outcompeting effortful corrections

6.4. Statistical Impact

• The generation effect averages about 0.4 standard deviations across 86 studies (Bertsch et al., 2007)
• In the original font-and-reasoning study, 90% of the clear-font group made at least one error against 35% of the hard-font group (Alter et al., 2007, N = 40)
• Pooled across the original and 16 replications, with more than 7,000 participants, the hard font had no effect (Meyer et al., 2015)
• The original disfluent-font memory study (N = 28) reported recall of about 86.5% versus 72.8%; the 2018 meta-analysis found no reliable benefit across text-learning studies
• Sans Forgetica showed no memory benefit in four experiments by Taylor et al. (2020), three preregistered experiments by Geller et al. (2020) and four experiments with nearly 300 participants by Huff et al. (2022)

7. The Hidden Benefits

In its reliable form, the processing difficulty effect is mostly good news.

• A free memory boost: Generating, explaining and self-testing cost nothing but effort, and they work across ages and subjects.

• A better signal than ease: Knowing the effect teaches people to stop treating fluency as evidence of learning, which fixes a lot of bad study habits.

• Tolerance for struggle: People who expect learning to feel hard are less likely to quit when it does. Elizabeth and Robert Bjork have written that learners often misread effortful learning as poor learning.

These benefits depend on the difficulty being solvable. A problem too hard to crack produces frustration and errors, not memory.

8. Self-Assessment: Do You Have This Bias?

8.1. Warning Signs Checklist

• You reread notes until they feel familiar and treat that feeling as readiness
• You prefer watching a worked solution to attempting the problem first
• You avoid self-testing because wrong answers feel discouraging
• You believe that harder-to-read text will help you or others remember it
• You highlight large parts of a text instead of summarising in your own words
• After a clear lecture or video, you feel sure you could explain the topic, but you have not tried
• You judge a course by how easy it was to get through
• You have adopted a study tip or "brain hack" without checking whether independent research supports it
• You give up on difficult material quickly because struggling feels like proof it is not working
• You rarely explain what you learned to anyone else

Scoring:

• 0-2 checked: Low susceptibility
• 3-5 checked: Moderate susceptibility
• 6-8 checked: High susceptibility
• 9-10 checked: Very high susceptibility

8.2. Self-Reflection Questions

1. When did you last remember something unusually well, and how did you learn it? Did you have to produce it, explain it or work it out?
2. Which of your study or work habits feel productive but leave you unable to recall much a week later?
3. How do you react when practice questions show you know less than you thought?
4. Have you passed on a learning tip because it sounded clever, without checking where it came from?
5. Where in your work could you swap a passive step (reading, watching) for an active one (answering, teaching, deciding)?

8.3. Quick Diagnostic Scenario

Scenario: You have a certification exam in two weeks, with a textbook, your notes and a bank of practice questions. A colleague recommends a font that is supposed to make studying more effective because it is hard to read.

How would you prepare?

• A) Convert your notes into the hard-to-read font and reread them several times, since harder must mean better → High susceptibility
• B) Reread the textbook, highlight key sections and try a few practice questions the night before → Moderate susceptibility
• C) Start with practice questions, check your answers, write short explanations of what you got wrong and space the sessions out → Low susceptibility

9. Identifying This Bias in Others

9.1. Behavioral Indicators

• Studies by rereading and highlighting rather than recalling or testing
• Treats a clear presentation as proof that the audience has learned the content
• Promotes hard fonts or similar surface tricks as memory aids
• Avoids practice tests because low scores feel demoralising

9.2. Conversational Red Flags

Phrases people say when under this bias:

• "I've read it three times, I know it."
• "That was so clear, everyone will remember it."
• "Put it in a weird font, it makes the brain work harder."
• "If it's hard, it must be working."

Types of arguments they make:

• Treating any difficulty as beneficial, whether or not it targets meaning
• Citing one striking study as settled science

Questions they avoid asking:

• "Can I recall this without looking?"
• "Has this been independently replicated?"

9.3. Situational Triggers

• Exam preparation under time pressure, when rereading feels safer than testing
• Polished training materials that are effortless to consume
• Viral articles and talks about learning "hacks"
• Settings that reward completion rather than recall

10. Cognitive Debiasing Strategies

10.1. Immediate Techniques

• Close the book and recall: After reading a section, write or say the main points without looking, then check
• Predict before you peek: Attempt a problem before seeing the solution
• Explain it in your own words: Teach the idea to someone, even an imaginary listener
• Ask "what kind of difficulty?": When something feels hard, check whether you are thinking about meaning or just decoding the page

10.2. Long-Term Strategies

• Build retrieval into routine: Make spaced practice questions or flashcards your default study method
• Test your confidence: A week later, quiz yourself on material you felt sure about
• Read the replications: Before adopting a learning technique, check whether large or preregistered studies support it

10.3. Environmental Design

• Readable materials, active tasks: Keep text legible and put the difficulty into questions, problems and discussion
• Quizzes over summaries: Build short retrieval questions into training modules and meetings
• Teach-back by default: In healthcare, onboarding and support, ask people to explain the key steps back
• Measure delayed recall: Judge training by what people remember after a week, not by satisfaction scores

10.4. When to Seek External Input

• When choosing a learning intervention for a school or company, especially one sold as a "hack"
• When a single striking study is the main evidence for a practice

Whom to ask: Cognitive or educational psychologists who know the replication literature. Ask: "Has this been tested independently, with a large sample, against a meaning-focused alternative?"

11. Practical Exercises

Exercise 1: Read Versus Generate

• Objective: Experience the generation effect directly and see how poorly ease predicts memory
• Time required: 30 minutes, plus 5 minutes two days later
• Materials needed: A list of 20 facts or word pairs, paper and pen
• Difficulty level: Beginner
• Instructions:
  1. Study half the list by reading each item twice
  2. Study the other half by covering part of each item and producing it from a cue, such as a first letter
  3. Rate how well you think you learned each half from 1 to 10
  4. Two days later, test yourself on both halves without looking
• Reflection questions:
  • Which half did you remember better?
  • How well did your 1-to-10 rating predict your recall?
• Frequency: Once, then again with new material if the result surprised you

Exercise 2: The Font Check

• Objective: Test whether perceptual difficulty helps your own memory
• Time required: 40 minutes over two sessions
• Materials needed: Three comparable short texts, a printer or screen
• Difficulty level: Intermediate
• Instructions:
  1. Set one text in a clear font and one in a hard-to-read font (small, grey or unusual)
  2. Read each for the same amount of time
  3. Read the third text in a clear font, stopping after each paragraph to summarise it in one sentence
  4. Two days later, write down everything you remember from each text
• Reflection questions:
  • Did the hard font change how much you remembered?
  • How did the summarising condition compare?
• Frequency: Once

Exercise 3: Teach-Back Week

• Objective: Turn passive learning into active generation
• Time required: 10 minutes per day for one week
• Materials needed: A notebook or voice recorder
• Difficulty level: Intermediate
• Instructions:
  1. Each day, pick one thing you learned from a meeting, article or class
  2. Without looking at sources, explain it as if to a newcomer, in a paragraph or a two-minute recording
  3. Check the source and mark what you got wrong or left out
  4. At the end of the week, explain all seven items again from memory
• Reflection questions:
  • Which items were hardest to explain on the first day?
  • Which ones survived to the end of the week?
• Frequency: One week, repeated monthly if useful

Daily Practice

The Two-Minute Recall

At the end of each day, spend two minutes writing down the main things you read, heard or learned, without looking anything up.

• Suggested duration: 2 minutes
• Best time of day: Evening
• How to track progress: Keep the notes in one file and occasionally check them against sources

Weekly Challenge

Replace One Passive Habit

Each week, pick one passive habit (rereading, watching a recorded talk, skimming a summary) and replace it with an active one (self-testing, explaining, working a problem first).

• Expected outcomes after 4 weeks: Better retention and a clearer sense of which methods only feel productive
• Journaling prompts for reflection:
  • Which replacement felt most uncomfortable?
  • What did I remember a week later that I would normally have forgotten?

12. For Specific Audiences

For Leaders and Managers

• Training Design: Build exercises, case discussions and short quizzes into training rather than relying on polished slides
• Meeting Follow-Through: Ask people to restate decisions in their own words before the meeting ends
• Evaluate Retention: Measure what employees remember weeks later, not how much they enjoyed the session

For Parents and Educators

• Age-Appropriate Explanation: "When your brain has to work a little to figure something out, you usually remember it better. But making the words blurry doesn't count. The work has to be thinking."
• Retrieval Practice: Use low-stakes quizzes and "write what you remember" exercises regularly
• Let Them Struggle a Bit: Give students a chance to attempt problems before showing solutions, keeping problems within reach
• Keep Materials Legible: The evidence does not support hard fonts as a learning aid

For Healthcare Professionals

• Teach-Back: Ask patients to explain instructions in their own words
• Clear Materials: Keep leaflets and discharge instructions easy to read; disfluency discourages adherence
• Active Learning: Use simulation, case questions and retrieval practice in continuing education

For Financial Professionals

• Client Engagement: Have clients work through scenarios, such as calculating the effect of fees over 20 years, rather than only reading summaries
• Fluency Caution: Easy-to-absorb pitch decks and names can inflate confidence in an investment
• Personal Learning: Write out your investment theses and models yourself to retain and test them

13. Interactions with Other Biases

Biases That Amplify This One

Biases That Counteract This One

Common Bias Chains

Illusion of Competence (Fluency Illusion) → Rereading → Overconfidence Effect → Poor exam performance

Rereading makes material feel familiar, so the student keeps rereading instead of self-testing, and confidence rises while recall stays flat.

Processing Difficulty Effect (misread) → Effort Justification → Confirmation Bias → Wasted teaching hours

A teacher switches worksheets to hard fonts, values the method because it costs effort, and then notices only the students who do well.

Interrupting the chain: Test recall a few days after studying, and look for large replications before adopting a technique. Both replace a feeling with evidence.

14. Cultural Perspectives

Most research on this effect comes from university students in North America, Europe, Australia and New Zealand, and there is little systematic cross-cultural work.

Systems that stress recitation and practice fit well with retrieval and generation, though rote recitation without understanding is still shallow processing.

Reading in a second language is a built-in form of disfluency. Boaz Keysar, Sayuri Hayakawa and Sun Gyu An (2012) reported that working through decision problems in a foreign language reduced some framing effects, which they attributed to more deliberate processing. Memory is a separate question; decoding unfamiliar words can crowd out attention to meaning.

How learners respond to difficulty, whether they persist or quit, is probably more culturally shaped than the memory effect itself.

15. Myths and Misconceptions

16. Expert Insights

Learning conditions that slow initial progress, such as spacing, interleaving and testing, often produce better long-term retention, but only when the learner can overcome the difficulty. — Robert A. Bjork, paraphrasing his "desirable difficulties" argument (1994)

"...no evidence of a disfluent font benefit under any conditions." — Andrew Meyer, Shane Frederick and colleagues, "Disfluent Fonts Don't Help People Solve Math Problems," Journal of Experimental Psychology: General 144(2) (2015), pp. e16–e30 https://doi.org/10.1037/xge0000049

"Our findings suggest that Sans Forgetica really is forgettable." — Jason Geller, Sara D. Davis and Daniel J. Peterson, "Sans Forgetica Is Not Desirable for Learning," Memory 28(8) (2020), pp. 957–967 https://doi.org/10.1080/09658211.2020.1797096

Experiences of difficulty may act as an alarm that switches on more analytic reasoning. — Alter, Oppenheimer, Epley and Eyre, 2007 (paraphrasing their proposal, which later large replications did not support for font-based difficulty)

17. Key Takeaways

1. Effort can improve memory: Information you work out, generate or connect yourself is often remembered better than information handed to you whole.

2. The kind of difficulty matters: Difficulty that engages meaning (generation, inference, retrieval) reliably helps. Difficulty that only makes text harder to see does not.

3. The font effect did not replicate: The 2007 reasoning result failed in a pooled analysis of more than 7,000 participants (Meyer et al., 2015), and a 2018 meta-analysis found no reliable learning benefit of perceptual disfluency.

4. Sans Forgetica is a cautionary tale: A font designed to be desirably difficult produced no memory benefit in independent tests.

5. Ease is a poor guide to learning: Fluent study feels productive and often produces weak retention; effortful study feels worse and often works better.

6. Desirable difficulties must be solvable: Struggle helps only when learners can overcome it.

7. Practical rule: Keep materials easy to read, and put the effort into recalling, explaining and applying what they say.

18. Further Resources

Academic Papers

• Slamecka, N. J., & Graf, P. (1978). The generation effect: Delineation of a phenomenon. Journal of Experimental Psychology: Human Learning and Memory, 4(6), 592-604. https://doi.org/10.1037/0278-7393.4.6.592
• O'Brien, E. J., & Myers, J. L. (1985). When comprehension difficulty improves memory for text. Journal of Experimental Psychology: Learning, Memory, and Cognition, 11(1), 12-21. https://doi.org/10.1037/0278-7393.11.1.12
• Bertsch, S., Pesta, B. J., Wiscott, R., & McDaniel, M. A. (2007). The generation effect: A meta-analytic review. Memory & Cognition, 35(2), 201-210. https://doi.org/10.3758/bf03193441
• Alter, A. L., Oppenheimer, D. M., Epley, N., & Eyre, R. N. (2007). Overcoming intuition: Metacognitive difficulty activates analytic reasoning. Journal of Experimental Psychology: General, 136(4), 569-576. https://doi.org/10.1037/0096-3445.136.4.569
• Diemand-Yauman, C., Oppenheimer, D. M., & Vaughan, E. B. (2011). Fortune favors the bold (and the italicized): Effects of disfluency on educational outcomes. Cognition, 118(1), 111-115. https://doi.org/10.1016/j.cognition.2010.09.012
• Meyer, A., Frederick, S., Burnham, T. C., Guevara Pinto, J. D., Boyer, T. W., Ball, L. J., Pennycook, G., Ackerman, R., Thompson, V. A., & Schuldt, J. P. (2015). Disfluent fonts don't help people solve math problems. Journal of Experimental Psychology: General, 144(2), e16-e30. https://doi.org/10.1037/xge0000049
• Rummer, R., Schweppe, J., & Schwede, A. (2016). Fortune is fickle: Null-effects of disfluency on learning outcomes. Metacognition and Learning, 11(1), 57-70. https://doi.org/10.1007/s11409-015-9151-5
• Eitel, A., & Kühl, T. (2016). Effects of disfluency and test expectancy on learning with text. Metacognition and Learning, 11(1), 107-121. https://doi.org/10.1007/s11409-015-9145-3
• Xie, H., Zhou, Z., & Liu, Q. (2018). Null effects of perceptual disfluency on learning outcomes in a text-based educational context: A meta-analysis. Educational Psychology Review, 30(3), 745-771. https://doi.org/10.1007/s10648-018-9442-x
• Taylor, A., Sanson, M., Burnell, R., Wade, K. A., & Garry, M. (2020). Disfluent difficulties are not desirable difficulties: The (lack of) effect of Sans Forgetica on memory. Memory, 28(7), 850-857. https://doi.org/10.1080/09658211.2020.1758726
• Geller, J., Davis, S. D., & Peterson, D. J. (2020). Sans Forgetica is not desirable for learning. Memory, 28(8), 957-967. https://doi.org/10.1080/09658211.2020.1797096
• Roediger, H. L., & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249-255. https://doi.org/10.1111/j.1467-9280.2006.01693.x

Books

• Brown, P. C., Roediger, H. L., & McDaniel, M. A. (2014). Make It Stick: The Science of Successful Learning. Belknap Press of Harvard University Press. https://doi.org/10.4159/9780674419377
• Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux.

Book Chapters

• Bjork, R. A. (1994). Memory and metamemory considerations in the training of human beings. In J. Metcalfe & A. P. Shimamura (Eds.), Metacognition: Knowing about Knowing (pp. 185-205). MIT Press. https://doi.org/10.7551/mitpress/4561.003.0011
• Bjork, E. L., & Bjork, R. 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.

19. Summary Card

20. Glossary of Terms Used

21. Discussion Questions

For book clubs, classrooms, or self-reflection:

1. Think of something you learned years ago and still remember clearly. How did you learn it? Was it effortful?

2. Why might a striking result like the hard-font study spread so quickly, while its failed replications get much less attention?

3. If effortful learning works better but feels worse, how should teachers and trainers respond when students say a method "isn't working"?

4. Where is the line between a desirable difficulty and one that only frustrates? How could you tell in your own learning?

5. As AI tools produce summaries and answers on demand, what happens to the kinds of effort that make information stick? How might we design learning to keep those efforts in place?

ਆਖਰੀ ਅੱਪਡੇਟ 4 ਨਵੰਬਰ 2026