Article: Processing Fluency and Clarity
Reviewed: November 2026 · Method version 2 · How these reviews work
Each factual claim in the article was checked against the source it cites, then challenged by reviewers whose job was to find the strongest evidence against it: replication attempts, meta-analyses, later papers that cite the original work, and research from other fields that study the same mechanism. A separate reviewer weighed both sides, and a script re-checked every quoted source. Where a claim was overstated or misattributed, the article was corrected. The research below supports the article as it now reads.
When Daniel Oppenheimer surveyed undergraduates, a majority admitted to deliberately making their vocabulary more complex to give the impression of intelligence.
Oppenheimer’s survey finding is echoed by Brown, Anicich and Galinsky (2020), who found that feeling low in status causally increases jargon use.
People rarely think about it directly. They notice the result, a sense that a text is clear or a slog, and unless something else obviously explains the difficulty, they tend to credit that feeling to the content and its author rather than to the typography or the word choice.
The misattribution of reading ease holds when no other cause is salient; when one is, readers correct for it (Oppenheimer 2006, Experiment 5).
A reader who struggles with your onboarding copy doesn’t think “this sentence has too many clauses.” They are more likely to conclude the product is complicated.
No study tests product judgments directly, but a 2026 experiment found that difficulty caused by jargon was misattributed to institutions, and a 2025 study found plain terms lowered the perceived difficulty of a financial decision.
Oppenheimer’s 2006 paper in Applied Cognitive Psychology ran five experiments; in the first three, more complex vocabulary led readers to judge the authors as less intelligent.
A 2021 survey experiment found business writers seemed more confident when they cut needless words and jargon; field data (Markowitz and Shulman 2021) show complexity can help where effort signals value.
Statistically, the penalty ran through fluency: complex wording made texts harder to process, and that difficulty lowered the rating.
In a related fluency effect, a 2018 meta-analysis found subjective ease explains about a third to half of the total effect, so partial mediation is the norm.
The fourth experiment removed vocabulary from the equation. Identical texts were set in an easy or a hard-to-read font, and the authors of the hard-to-read versions were judged less intelligent. So the experience of reading matters, not just the words. Font effects on their own have proven less reliable in later studies of other judgments, so the wording is the safer place to start.
Font-only fluency manipulations have failed to replicate for truth judgments (JESP 2025), moral judgments (Cognition 2020) and reasoning (Meyer et al. 2015); no direct replication of Oppenheimer’s Experiment 4 was found.
You can’t count on that rescue in an interface. Users rarely have an innocent explanation for why your settings page is hard to read, so they are likely to blame the product.
Applying Experiment 5 to interfaces is an inference; a 2026 experiment shows difficulty can be misattributed to institutions.
In Song and Schwarz’s study, students who read exercise instructions in ornate Brush font rather than Arial expected the routine to take longer and feel harder, and were less willing to make it part of their day.
Song and Schwarz used small student samples; later font-disfluency effects on other outcomes have been fragile, though a 2017 review of 40 fluency studies supports the effort-prediction link.
When instructions are hard to read, the task itself looks hard to do.
The link between reading ease and perceived task difficulty extends beyond fonts to wording (BMC Research Notes 2017; Journal of Public Policy and Marketing 2025).
Nothing about the task changed. Numbered steps and readable type make it look like a short job, and how hard a task looks shapes whether people are willing to start it.
Song and Schwarz found that readable instructions raised willingness to act; numbered steps and contrast were not tested.
In error messages, every term users have to decode makes the fix seem further away.
Experimental work on error messages (CHI 2019) and plain terminology (2025) is consistent with this advice.
Legalistic phrasing makes an easy exit sound like a negotiation; plain phrasing makes a generous policy read as generous.
Plain-language redrafting of contracts improves comprehension (Masson and Waldron 1994), though comprehension stayed low overall.
Use the plain word your customer would use. For expert readers, keep the technical terms they use themselves and cut the rest.
NN/g (2017) recommends plain, succinct writing for experts but keeps shared technical terms; Oppenheimer tested needless long words, not domain vocabulary.
Fluency is one of the first things a message has to get past. Readers sense quickly whether reading will be work, and that feeling can color how they judge the product, its instructions and whoever wrote them.
Fluency effects on truth and moral judgments have been small or failed to replicate in registered and multi-study tests (2020, 2023, 2025), so fluency is one influence on judgment, not a gate.
Readers hear differences in tone between equally readable sentences at once and use them to decide whether the company deserves their confidence.
Readers infer writer confidence from style (Business and Professional Communication Quarterly 2021).
These edits were made on November 12, 2026 after the review.
Claim reviewed: People rarely notice processing fluency directly; they notice a sense that a text is clear or a slog and attribute that feeling to the content and its author rather than to the typography or word choice.
Now reads: People rarely think about it directly. They notice the result, a sense that a text is clear or a slog, and unless something else obviously explains the difficulty, they tend to credit that feeling to the content and its author rather than to the typography or the word choice.
Why: The paper does show that how easy a text is to process shapes how its author is judged. But the same paper’s fifth experiment shows people stop doing this when the cause of difficulty is obvious, so the misattribution is conditional, and no study cited tests that people ‘rarely notice’ fluency. Later research also finds fluency explains only part of related judgment effects.
Claim reviewed: A reader who struggles with onboarding copy does not blame the sentence; they think the product is complicated.
Now reads: A reader who struggles with your onboarding copy doesn’t think “this sentence has too many clauses.” They are more likely to conclude the product is complicated.
Why: No source is cited, and studied judgments of authors, not products. Newer experiments in other fields do show that reading difficulty gets blamed on things beyond the writer: legal jargon lowered trust in the court system, and plain fee terms lowered perceived difficulty of investing. The extension is reasonable but should read as a likely outcome, not a finding. Now cites Journal of Applied Communication Research 2026 (legal jargon experiment).
Claim reviewed: In the fourth experiment identical texts set in a hard-to-read font led readers to judge the authors less intelligent, so the effect is about the experience of reading, whether the difficulty comes from the words or the display.
Now reads: The fourth experiment removed vocabulary from the equation. Identical texts were set in an easy or a hard-to-read font, and the authors of the hard-to-read versions were judged less intelligent. So the experience of reading matters, not just the words. Font effects on their own have proven less reliable in later studies of other judgments, so the wording is the safer place to start.
Why: Experiment 4 of found exactly this. But effects of font and contrast alone have been fragile in later research: several replication attempts on other judgments, including truth and moral judgments and reasoning, found no effect. The generalization is fair, but one experiment is a thin base for ‘the effect is about the experience of reading’ stated as settled.
Claim reviewed: Users rarely have an innocent explanation for why a settings page is hard to read, so they blame the product.
Now reads: You can’t count on that rescue in an interface. Users rarely have an innocent explanation for why your settings page is hard to read, so they are likely to blame the product.
Why: This is the article’s own reasoning from Experiment 5 of, not a tested result; no study checks whether users blame products for hard-to-read interface text. Related work supports misattribution to institutions, though jargon’s effect on credibility has been conditional in some studies. The sentence should read as a likely outcome.
Claim reviewed: Numbered steps and readable (full-contrast) type make a setup task look like a three-minute job, which is how a user decides whether to start it now or later.
Now reads: Nothing about the task changed. Numbered steps and readable type make it look like a short job, and how hard a task looks shapes whether people are willing to start it.
Why: The idea that readable instructions make a task look easier comes from Song and Schwarz, which tested fonts. No study found by pass 3 tests numbered steps or contrast on whether people start a task now or later, and ‘three-minute job’ is illustrative. The sentence is reasonable if tied to and stated less precisely. Now cites Song and Schwarz, Psychological Science (2008).
Claim reviewed: Use the word your customer would say on the phone, even in copy for experts; swapping plain words for longer ones to sound authoritative backfires, and expert buyers do not need the longer words.
Now reads: Use the plain word your customer would use. For expert readers, keep the technical terms they use themselves and cut the rest.
Why: Avoiding long words chosen to sound authoritative is well supported by. But the cited NN/g article itself says jargon helps when specialist readers share it, and other studies find technical language can help in some settings, so ‘the word your customer would say on the phone, even in copy for experts’ goes further than the sources.
Claim reviewed: Fluency is the first gate a message passes through: before readers weigh an argument they decide whether reading it will be work, and they hold that judgment against the product, its instructions and whoever wrote them.
Now reads: Fluency is one of the first things a message has to get past. Readers sense quickly whether reading will be work, and that feeling can color how they judge the product, its instructions and whoever wrote them.
Why: The sources support that reading ease affects how readers judge authors and tasks. But the claim that readers first decide whether reading will be work, before weighing an argument, is not tested, and several fluency effects on judgments have proven small or failed to replicate. It works as a summary if softened.