The study of text readability began in the 1920s when educators and researchers sought objective methods to match reading materials with student abilities. In 1923, Bertha A. Lively and Sidney L. Pressey published the first readability formula, which measured vocabulary difficulty based on word frequency to determine text difficulty. This groundbreaking work established that certain linguistic features could predict how easily a reader could comprehend a passage. [Wikipedia, Readability]
In 1935, William S. Gray and Bernice Leary published "What Makes a Book Readable," a comprehensive study that analyzed over 80 potential predictors of readability. Their research identified sentence length and vocabulary difficulty as the two strongest indicators, forming the foundation for nearly all subsequent readability formulas. [Wikipedia, William S. Gray]
Rudolf Flesch, an Austrian-born author and readability expert who earned a PhD in Library Science from Columbia University, published the Flesch Reading Ease formula in 1948 in the Journal of Applied Psychology. This formula, which calculates readability based on average sentence length and average syllables per word, became the most widely used readability metric in history. Flesch later collaborated with J. Peter Kincaid to develop the Flesch-Kincaid Grade Level formula for the United States Navy in 1975, which remains a standard for technical documentation and legal writing. [Wikipedia, Rudolf Flesch] [Wikipedia, Flesch-Kincaid Tests]
Robert Gunning developed the Gunning Fog Index in 1952, specifically designed for business writing and journalism. The formula calculates the years of formal education needed to understand a text on first reading, making it particularly popular among newspaper editors and corporate communications departments.
In 1975, Meri Coleman and T. L. Liau introduced the Coleman-Liau Index, which distinguished itself by relying on character count rather than syllable counting. This approach proved valuable for computer-based analysis since character counting is more reliable than automatic syllable detection. [Wikipedia, Coleman-Liau Index]
Harry McLaughlin developed the SMOG (Simple Measure of Gobbledygook) Index in 1969 specifically for healthcare and public health communications. McLaughlin designed this formula to be easily calculated by hand, making it accessible to healthcare providers creating patient education materials. The SMOG Index is still recommended by the National Institutes of Health and the American Medical Association for health-related documents. [Wikipedia, SMOG]
With the rise of computers and digital text processing, automated readability formulas have become integrated into word processors, content management systems, and writing assistance tools. Microsoft Word includes Flesch-Kincaid scores in its proofing tools. The Automated Readability Index (ARI), developed in 1967 for the US Air Force at Wright-Patterson Air Force Base, gained popularity with computer systems because its character counting approach was easier for software to implement reliably. [Wikipedia, Automated Readability Index]
Today, natural language processing and machine learning are producing more sophisticated readability models that consider factors beyond sentence length and word difficulty, including discourse structure, coherence, and background knowledge requirements. [Wikipedia, Natural Language Processing]
Every readability formula uses sentence length as a primary variable. Longer sentences require readers to hold more information in working memory, track more complex grammatical relationships, and parse more clauses before reaching the verb. Research consistently shows that as average sentence length increases, comprehension decreases, especially for less skilled readers. [Wikipedia, Sentence Clause Length]
Word difficulty is the second essential component of readability formulas. Different formulas measure word difficulty in different ways. Syllable based formulas like Flesch-Kincaid and Gunning Fog count syllables per word, assuming that longer words with more syllables are harder to read. Character based formulas like Coleman-Liau and the Automated Readability Index count characters per word, which correlates strongly with syllable count but is easier for computers to calculate accurately.
Research has shown that word frequency is actually a better predictor of difficulty than syllable or character counts. Common words like "the," "and," and "of" are recognized instantly by fluent readers, while rare words like "ubiquitous" or "esoteric" require more cognitive effort regardless of their length.
Readability formulas are linear regression equations derived from empirical studies. Researchers would give readers texts of varying difficulty, measure comprehension through tests, and then find the mathematical combination of variables that best predicted the test scores. The resulting formulas are weighted sums of the selected variables, typically with different coefficients for each variable based on its importance in predicting readability.
For example, the Flesch Reading Ease formula is: 206.835 - (1.015 x average sentence length) - (84.6 x average syllables per word). The constants in this equation were determined by analyzing the relationship between these text features and actual reading comprehension scores from thousands of readers.
Readability formulas have well known limitations. They do not account for reader background knowledge, motivation, or interest in the subject matter. A text about baseball might be easy for a baseball fan to read even if it scores poorly on a formula, while a formula suggests a text is easy but the reader lacks necessary vocabulary or concepts will still struggle.
Formulas also cannot measure text cohesion, the logical connections between ideas, or discourse structure. Two texts with identical sentence lengths and word difficulties might have very different actual readability if one presents ideas in a clear, logical order while the other jumps randomly between topics. For these reasons, readability formulas should be used as guidelines rather than absolute measures of text quality. [Wikipedia, Cohesion]
Developed: 1975 by Rudolf Flesch and J. Peter Kincaid | Variables: Sentence length, syllables per word
The Flesch-Kincaid Grade Level formula was created for the United States Navy to ensure technical manuals were written at appropriate reading levels. The formula outputs a U.S. school grade number. For example, a score of 8.0 means an eighth grader can understand the text. The formula is: 0.39 x (words per sentence) + 11.8 x (syllables per word) - 15.59. This formula is built into Microsoft Word and is widely used for legal documents, technical writing, and educational materials. [Wikipedia, Flesch-Kincaid Grade Level]
Developed: 1942 by Rudolf Flesch | Variables: Sentence length, syllables per word | Scale: 0 to 100
The original Flesch formula measures readability on a scale from 0 to 100, where higher scores indicate easier text. Scores between 90 and 100 are considered very easy to read (typical for comic books), 60 to 70 is plain English (newspaper level), and 0 to 30 is very difficult (academic journals). This formula became the standard for assessing consumer documents, insurance policies, and government communications. Many countries require certain legal documents to achieve minimum Flesch Reading Ease scores. [Wikipedia, Flesch Reading Ease]
Developed: 1952 by Robert Gunning | Variables: Sentence length, complex words (three or more syllables)
The Gunning Fog Index estimates the years of formal education a person needs to understand a text on first reading. The formula counts words with three or more syllables as complex words, excluding proper nouns, familiar jargon, and compound words. The index is calculated as: 0.4 x (average sentence length + percentage of complex words). A score of 7 to 8 is considered ideal for business writing, while scores above 12 indicate texts that may be too difficult for general audiences. [Wikipedia, Gunning Fog Index]
Developed: 1975 by Meri Coleman and T. L. Liau | Variables: Characters per word, words per sentence
The Coleman-Liau Index differs from other formulas by using character count rather than syllable count. This makes it more reliable for computer analysis since counting characters is unambiguous while syllable counting can be difficult to automate. The formula is: 5.89 x (characters per word) - 0.3 x (words per sentence) - 15.8. The result corresponds to a U.S. grade level. The index is particularly useful for analyzing texts where syllable counting might be unreliable, such as technical documents with many abbreviations or specialized terms. [Wikipedia, Coleman-Liau Index]
Developed: 1969 by Harry McLaughlin | Variables: Words with three or more syllables in 30 sentences
The SMOG (Simple Measure of Gobbledygook) Index was designed specifically for health communications and is recommended by the National Institutes of Health. Unlike other formulas that average across entire texts, SMOG analyzes a sample of 30 sentences. The formula is: 1.043 x square root of (number of complex words x 30 divided by number of sentences) + 3.1291. SMOG typically gives higher grade level estimates than other formulas, making it a conservative measure often used for patient education materials. The American Medical Association recommends that patient information be written at or below a 6th grade SMOG level. [Wikipedia, SMOG Index]
Developed: 1967 for the US Air Force | Variables: Characters per word, words per sentence
The Automated Readability Index (ARI) was developed in 1967 by R. J. Senter and E. A. Smith at the Aerospace Medical Research Laboratories at Wright-Patterson Air Force Base for the US Air Force, to assess technical manuals and training materials. Like the Coleman-Liau Index, ARI uses character counts rather than syllable counts, which made it practical for early computer systems. The formula is: 4.71 x (characters per word) + 0.5 x (words per sentence) - 21.43. The ARI output corresponds to a U.S. grade level. One advantage of ARI is that character counting is unambiguous and can be performed by simple computer programs without complex linguistic analysis. [Wikipedia, Automated Readability Index]
Each readability formula has strengths and weaknesses. For general business and technical writing, Flesch-Kincaid Grade Level is widely accepted. For healthcare communications, SMOG is recommended by medical authorities. For newspaper and journalism, Gunning Fog Index is traditional. For automated systems where syllable counting is unreliable, Coleman-Liau or ARI are better choices. For best results, it is wise to compare multiple formulas to get a comprehensive understanding of text difficulty. [Wikipedia, Comparison of Readability Formulas]