The Effect of Disfluency on Retention. Why You May Want to Be A Bit Confusing in 2025 ?

A Story

It could have been a typical Wednesday afternoon, in 2011, in the university's psychology department. Connor Diemand-Yauman could have sat down across from Daniel Oppenheimer, a look of intrigue on his face.

"The data is in," he could have said, sliding a stack of papers across the table, "obfuscation predicts retention."

Oppenheimer's eyes could have lit up as he began to scan the pages.

"You're telling me the students who read the texts in difficult-to-read fonts performed ..."

... significantly better than those who read the same texts in easy-to-read fonts," Diemand-Yauman could have finished, a hint of a smile on his lips.

"This could be big," he could have said, his voice barely above a whisper, his eyes sparkling with excitement. "We could be looking at a simple, yet effective way to improve student outcomes, without breaking the bank."

Oppenheimer could have leaned back in his chair, put his arms up behind his head, and smiled. He could have said "we did it big guy. We proved the disfluency effect! This could be a fundamental shift in the way we think about learning."

The fog outside could have started to lift. A sense of excitement could have begun to build in the air. A new era in education could have begun.

Okay. Clearly we're dramatizing. The story we're telling most certainly didn't happen like that. But the questions is 'why dramatize?', and 'what are we dramatizing about?'

Oh ya ... and what's going on with all of these crazy fonts!

The Disfluency Effect

Okay, so now that I've got your attention. Let's ease up on the fonts for a bit. Here's what we know about what actually happened.

Back in 2011, Connor Diemand-Yauman and Daniel Oppenheimer published a paper titled "Fortune Favors the Bold (and the Italicized): Effects of disfluency on education outcomes."

They asked 222 students to read a passage of text. Some of them were given text printed in 'functional fonts', that are easy to read, others were given the same text in slightly obfuscated fonts, called now sometimes 'disfluent' fonts, because they are slightly more challenging to read.

Then they tested the students on comprehension and retention.

The students who read the more difficult to read text scored better on the tests. They comprehended more. They recalled more information. They learned more.

The final result of the study was that disfluency in design, or slight obfuscation, predicts retention, where 'predicts' is a reference to a standard OLS regression analysis, and retention refers to human memory.

Why

Diemand-Yauman and Oppenheimer suggest disfluent fonts increase cognitive processing effort. When text is presented in a difficult-to-read font, readers are more likely to engage in deeper cognitive processing, which can lead to better memory retention and understanding.

They suggest disfluent fonts make the information seem more challenging to learn. As a result, readers are more likely to concentrate carefully on the material, and engage more thoroughly with the content.

The subjective experience of difficulty associated with cognitive operations, known as disfluency, prompts readers to adopt more systematic and elaborative reasoning strategies rather than relying on heuristics or intuitive processing.

Additionally, some people believe disfluency can serve as a meta-cognitive cue, indicating to the reader that they may not have fully mastered the material, which again, can prompt them to increase effort.

Having experimented somewhat with design, text, and comprehension, in the context of the neuro-cognitive process and the emotional transportation associated with various ‘states’ of the ‘reading’ experience, here at Build Hello, we have an additional theory that elaborates on the possible causes of the ‘disfluency effect’.

In our theory, in the Deimand-Yauman and Oppenheimer study, the interaction between the reader and the disfluent font could have been characterized as an interaction between a human and a game, like a crossword puzzle.

When humans solve puzzles or complete challenges their brain releases dopamine, which motivates them to continue or repeat that activity.

Dopamine, the “reward drug”, provides a boost of pleasure and contributes to happiness, enhanced cognitive ability, and improved ability to learn and pay attention.

While increased focus and effort is a likely contributor to the disfluency effect of fonts on retention, the ‘puzzle effect’, as we call it, could be as impactfull on educational outcomes.

An additional paper, on the subject of math and games, by Haase & Hanel in 2022, titled “Priming creativity: Doing math reduces creativity and happiness whereas playing short online games enhance them”, could suggest some guidelines in terms of how much obfuscation is good obfuscation if one aims to trigger the disfluency effect.

Haase & Hanel suggest math tasks reduce creativity while game play enhances it. More precisely they say:

Your subscription could not be saved. Please try again.

Your subscription has been successful.

"Games can elicit an interactive, flexible, broad-associative mindset that facilitates creative thinking and problem-solving. In contrast, math tasks requiring a high concentration level tend to harm subsequent creative performance. Designers of digital learning environments should keep these connections in mind."

~ Haase & Hanel, 2022

Too much obfuscation could trigger the same effects as mathematics on creativity, negatively impacting cognitive retention.

We should say at this point, in our experience, complex mathematics is good for brain growth, what we are talking about here is font use on text.

If you’re a math teacher, please don’t read Haase & Hanel, and remove math from your syllabus!

The disfluency effect, though it has many critics, one of which we are going to mention, is interesting enough to the field of design, as well as other fields, like curriculum writing, and marketing, that we’d like to share some materials where you can apply it, and some fonts you can apply it with.

Materials You Might Consider Ripe for Disfluency

Digital and in-person classroom materials: test your disfluent fonts on classroom materials, such as textbooks, presentations, handouts, and worksheets
Educational videos: subtitles or captions could be written in mildly disfluent fonts
Marketing brochures: experiment with disfluent fonts in marketing brochures, flyers, business cards, posters, and other promotional materials
Website content: incorporate disfluent fonts into website content, such as blog posts or product descriptions
Social media posts: use disfluent fonts in social media posts
Online Quizzes and tests: tryout disfluent fonts in online quizzes and tests
Infographics: infographics, which are generally already attractive to readers, could be made more attractive with disfluency fonts

Some Fonts You Might Want to Experiment With When Seeking Disfluency

First of all, take note of the font used in the original Diemand-Yaumand and Oppenheimer study: Comic Sans MS.

The obfuscation and disfluency in Comic Sans MS is subtle. The font is slightly bulging, slightly ‘display’, mildly cursive, and comical.

The important notion is that subtle disfluency has an effect. In general, when applying disfluency, subtlety should not be overlooked.

The fonts we’re sharing in this article are, for the most part, overtly disfluent – one could say they are ‘over the top’.

We’re illustrating the point!