What to know about plain language, clear communications, accessibility, and “readability” tools

This blog has roots in recurring questions and comments about plain language and using readability tools. See this blog for more about:

• key “readability” tools used now
• some history
• why the key tools don’t stand up to scrutiny
• what to use instead

In posts, presentations, and conference conversations, I have said that “readability” tools such as readability scores and reading grade levels are not effective for looking at plain language, clear communications, and accessibility.

Leaders in the field of plain language, clear communications, and accessibility offer varied research and guidance on this subject. These include the following:

• Iva Cheung and other members I worked with on the Accessibility Standards Canada (ASC) Technical Committee for Plain Language National Standard of Canada (CAN-ASC-3.1:2025)
• Ginny Redish, Karen Schriver, Gael Spivak, and others involved with developing and promoting the International Organization Standardization, Plain Language, Part 1: Governing Principles and Guidelines (ISO 24495-1:2023)
• members of PLAIN (Plain Language Association International), the International Plain Language Federation (IPLF), and Editors Canada, among others

I get into details later in this blog and offer references as well as my own experience with plain language, clear communications, and accessibility.

But the key point is what to use instead of “readability” tools to get the results that governments, businesses, and other organizations and their audiences want.

What are the key “readability” tools now?

Today, most writers, editors, other folks in communications, and decision-makers will think of these tools shown in Microsoft Word’s “Readability Statistics” as the key tools:

1. the Flesch Reading Ease, and
2. the Flesch-Kincaid Grade Level.

Microsoft Word and these tools are ubiquitous and widely adopted.

The Flesch Reading Ease and the Flesch-Kincaid Grade Level do the following:

• focus on blocks of running text
• count the average sentence length (number of words divided by the number of sentences)
• count the average number of syllables per word (number of syllables divided by the number of words)
• apply a formula to the counts

The Flesch Reading Ease uses a scale to 100.

The Flesch-Kincaid Grade Level is said to approximate grade levels in the United States. According to Microsoft, “a score of 8.0 means that an eighth grader can understand the document.” (See Microsoft, “Get Your Document’s Readability and Level Statistics in Microsoft Word”.)

What is the history of “readability” tools?

Here’s a slice of the history from my perspective. (Other plain language specialists will have different experiences depending on how long they have been doing this work, where, their training, and so on.)

In the 1980s, I studied many ways to think about reading in English, writing, editing, design, publishing, books, magazines, how readers learn to read, what can make texts easier to read, what can make texts harder to read, how readers can increase their skills and confidence, and writing and editing for different audiences—from children to adults.

Some of the above introduced me to readability tools, including the Flesch-Kincaid tools, the Gunning Fog Index, the Dale-Chall, and the SMOG Index (with the last short for “simple measure of gobbledygook”).

Here’s what I saw among readability tools in the 1980s:

• Many tools set out to measure how readable text was for adults.
• Most were developed in the United States.
• Most focused on English.
• Many were developed from the 1940s to 1970s.
• Some were developed for the military or marketers to use.
• Most depended on counting sentences, words, and syllables.
• Most focused on blocks of narrative text or “running text” but not headings, vertical lists, and visuals.
• Only some factored in words that were categorized as familiar to an audience.
• Most assumed standardized teaching, learning, and reading across the United States—not variations across regions, not variations for specialized areas of study or work, and not other countries.
• They focused on reading words. They didn’t focus on understanding the words and using that understanding to do a task.

Since the 1980s, more tools have been developed or updated. For example, Readable has a handy summary of many tools. Dates given for Flesch and Kincaid vary, but their work builds on Flesch’s thesis in the 1940s. (See Oxford University Press blog, “Reading, writing and readability, appreciating Rudolph Flesch.”)

Why are the key “readability” tools used now not effective?

Focusing on the Flesch-Kincaid tools seems appealingly simple, indisputable, and applicable. After all, we can conveniently count sentences, words, and syllables in a Microsoft Word document. So, we can easily confirm readability! Right?

Wrong.

See later in this blog for research, guidelines, and standards. But first let’s consider an example.

A sample to scrutinize

Suppose a sample text scores an 8.0 with the Flesch-Kincaid Grade Level.

• Does that score mean that the average Grade 8 student in the United States can read, understand, and use the text?
• What does that mean for audiences in countries other than the United States and in languages other than English?
• What does that mean for audiences of adults instead of Grade 8 students, who are about 13 or 14 years old?
• How are differences for adults—differences in needs, knowledge, and lived experience—addressed by that metric?
• When adult audiences are described as having a Grade 8 reading level, what are possible impacts on word choice, tone, and trust?
• What does that score mean for adults and children who have disabilities or face barriers to access and inclusion?
• What does that score based on research last century mean for audiences now?

Equally, what does that score of 8.0 mean for governments, businesses, and other organizations trying to develop communications that are plain, clear, accessible, accurate, trusted, and useful?

The assumptions and gaps

Using Flesch-Kincaid tools for plain language assumes:

• Plain language is all about reading text.
• Text with more words per sentence and more syllables per word is more complex and less plain.
• Plainer language requires shorter sentences and words of fewer syllables.

But if we count sentences, words, and syllables—and make changes based on the counts—we still do not know about understanding, accuracy, and usefulness. In fact, we could “game” the score to get a required Flesch-Kincaid Grade Level but reduce clarity, trust, and impact.

I could demonstrate word changes to change reading scores (for example, in public health messages, emergency instructions, or legal terms and conditions). But my point is that the Flesch-Kincaid tools do not stand up to what we know now about reading, readability, plain language, clear communications, and accessibility.

What leaders in the field say about “readability” tools

Ginny Redish

Ginny Redish and others have been involved with developing and promoting the International Organization for Standardization, Plain Language, Part 1: Governing Principles and Guidelines (ISO 24495-1:2023).

In their 2019 article, “Readability Formulas: Seven Reasons to Avoid Them and What to Do Instead,” Ginny Redish and Caroline Jarrett set out both reasons and examples from decades of experience and research.

The reasons to avoid readability formulas include the following:

• Readability formulas are not reliable.
• The scores are not valid.
• The grade levels are meaningless for adults.
• Formulas don’t consider the meaning of words.

For more reasons, examples, and relevant research, see the 2019 article.

Karen Schriver

Another leader in the field, Karen Schriver,  has pointed out that the study and practice of plain language has developed from focusing on readability at the word and sentence level to a broader view of usability. (See Schriver, “Plain Language in the US Gains Momentum: 1940-2015” in IEEE Transaction on Professional Communication, vol 60, issue 4, December 2017.)

International Plain Language Federation, PLAIN, and Accessibility Standards Canada

This change from focusing on the word and sentence to a broader view is reflected in the multiple aspects of communication now evident in the definition of plain language.

The International Plain Language Federation (IPLF) defines plain language as follows:

A communication is in plain language if its wording, structure, and design are so clear that the intended readers can easily find what they need, understand what they find, and use that information.

The IPLF definition gives weight to structure and design as well as wording, and to what the intended readers (or audience) needs, understands, and will use. This definition is now used in the following standards, which are both based on extensive research, experience, and best practices:

• International Organization for Standardization’s Plain Language, Part 1: Governing Principles and Guidelines (ISO 24495-1:2023)
• Canada’s Plain Language, National Standard of Canada (CAN-ASC-3.1:2025), but with a key change from “readers” to “audience.”

Gael Spivak

Gael Spivak is a Canadian plain language specialist, editor, and federal public servant who now chairs a committee of the International Plain Language Federation. That committee helps countries around the world to translate, localize, adopt, and implement the International Organization for Standardization (ISO) standard.

Gael wrote the post “Readability Formulas, Programs and Tools: Do They Work for Plain Language?” for Canada’s federal government blog called Our Languages. Some key points I would like to highlight are the following:

• Readability tools do not effectively evaluate what readers read and understand.
• Readability tools do not focus on what plain language specialists now know matters to achieve plain language (such as using words familiar to the intended audience, chunking content, and laying out text with white space).
• The ISO Plain Language, Part 1: Governing Principles and Guidelines does not support using readability formulas.
• The ISO Plain Language, Part 1 reflects expertise from around the world and in multiple languages—a key consideration in the face of the English-bias of the Flesch-Kincaid tools.

Iva Cheung

Iva Cheung is a plain language specialist, scholar, editor, trainer, researcher, and more.

As Iva says, “There are a lot of misconceptions. When I tell people I do plain language work, sometimes they think that it’s all about writing short sentences with short words or writing to a certain grade level…. Unfortunately, things are not quite so simple.” (See Grammar Girl episode, “The Case for Language Clarity, with Iva Cheung.”)

From 2020 to 2025, Iva served on the Accessibility Standards Canada (ASC) Technical Committee for Canada’s Plain Language, National Standard of Canada (CAN-ASC-3.1:2025). The standard was developed by Iva, myself, and other plain language and accessibility specialists across Canada, in English and French.

The standard’s clause 11.1 cautions:

“Readability formulas and reading levels are unreliable and ineffective ways to assess whether your communication is easy to read, understand, and use.”

And critically, Iva says:

[the intended audience is] “the most important part of the definition of plain language. … Something can be plain to one audience and not plain to another. And the only way you can tell if something is plain is if you ask your intended audience, which means that user testing is a core part of the plain language process.”

(See Grammar Girl episode, “The Case for Language Clarity, with Iva Cheung.”)

Final thoughts on readability tools and what to use instead

From my own experience as a writer, editor, and plain language specialist in Canada and working over decades, I have seen and used multiple ways to learn from readers and other audiences about what they need and want, understand, and use. Working with others, we have pilot tested, reviewed, asked for feedback, revised, updated, and started fresh.

Over time, I have seen some changes in practice:

• first expert reviews and user testing
• then writing and editing to meet a target reading grade level
• then using readability tools with caution (for example, “try for ‘a Grade 8’ but highlight key terms and define them”)
• then more evidence that “readability” tools are flawed
• now increasing use of user testing and reviews by subject experts and plain language experts

What audiences want, need, know already, need to know, read, understand, and use varies hugely. As Iva Cheung says, “Something can be plain to one audience and not plain to another.” And how audiences access information and the barriers they face vary.

So, what makes plain language and clear, accessible communications often means the following:

• using the words that are familiar to the audience (even if they have more syllables than other options)
• identifying and defining key but unfamiliar technical terms if the audience must know and use them (even if that increases “readability” scores)
• adding meaningful headings to help the audience skim, scan, and understand the structure
• adding supportive visuals
• structuring lists
• changing formats, structures, wording, and design to address audience use and feedback

In conclusion, “readability” tools are not effective for plain language. But standards such as Canada’s Plain Language, National Standard of Canada (CAN-ASC-3.1:2025) offer guidance, support, and extensive examples.

Thanks and some history

For this blog, I combined a series of 10 posts I wrote and published on LinkedIn in late June 2026, then edited the text further.

Thanks to Jenny O’Reilly for a related post on LinkedIn. Jenny’s very wise post and some repeated questions, comments, and assertions from others before Jenny’s post urged me to get my thoughts down on paper.

Thanks to Iva Cheung, Gael Spivak, Ginny Redish, and Karen Schriver for their comments on my draft. Stan Skinner, director of Centennial College’s Book Editing and Design program, first introduced me to readability tools.