Abram (1981) states that the history of readability dates back to 900 A.D. when word counts were used as a rough index of reading ease. Lorge (1944) explains how in compiling and studying the body of laws called the Talmud, the Talmudists counted the occurrences of words and ideas in seeking to distinguish differences in meaning. Modern research into readability began in 1921 when Thorndike published a list of English words used more frequently in texts. Assumptions were made that the more frequently a word was used, the more familiar readers became with it and the easier it was to read. During the 1920's, research activities concentrated on looking for word factors that could be used to predict readability. Research broadened during the 1930's and throughout the 1940's, deriving formulas that could accurately predict readability using the least number of factors.

A readability formula or index "is like a yardstick that helps us measure certain qualities in the writing so we can make objective judgements about reading level" (Laubach & Koschnick, 1977, p. 12). Many readability formulas have been developed as a result of research into factors within writing that correlated highly with style difficulty. Most readability formula values are calculated by measuring sentence and word familiarity or word length. There are several widely used readability formulas.

The Lorge Formula. Lorge first published his formula for children's material in 1939. The formula was designed to cover grades 3-12. It set the stage for many to follow by using the Standard Test Lessons in Reading (McCall & Crabbs, 1925) as the criterion of difficulty. The formula considers three factors: average sentence length in words, number of prepositional phrases per 100 words, and the number of different hard words on the Dale list of 763 words.

The Flesch Formula. In 1943, Flesch published his formula using the original McCall-Crabbs Standard Test Lessons in Reading as a criterion. In designing it for general adult reading matter, he felt it gave proper attention to abstract words as well as sentence length. The formula uses four factors: number of syllables per 100 words, average number of words per sentence, number of personal words per 100 words, and number of personal sentences per 100 sentences.

Dale-Chall Formula. Similar to Flesch's, it used the 1925 McCall-Crabbs Standard Test Lessons in Reading as a criterion. It has a 3,000 word list, which was deemed preferable to the 763 word list used by Lorge and especially for the more difficult levels of readability. Flesch's count of personal references was avoided as unnecessary, and only two factors were used in order to keep it easy to use. They are the average sentence length and the percentage of unfamiliar words (words outside the Dale list of 3,000 words).

The Fog Index. Gunning's Fog Index (1952) is similar to Flesch's Reading Ease formula. Rather than counting syllables as Flesch did, Gunning proposed counting words of three or more syllables. He termed these "hard words". The formula is based on two counts: average sentence length and percentage of words having three or more syllables.

The Readability Graph. Fry developed a "Readability Graph" in 1965 for predicting readability. He used the common formula variables of syllables per 100 words and words per sentence. The user marks the counts of the variables on a graph and then reads the readability grade score directly from it. Fry's graph has been validated using both primary and secondary level materials.

The Bormuth Formula. In 1969 Bormuth published the most extensive readability analysis yet made. Bormuth used 330 passages which were about 100 words each and which ranged in difficulty from first grade to college and covered a wide range of subject matter. The cloze procedure of deleting every fifth word was used as the criterion of difficulty. This formula contained from 14 to 20 variables each. Bormuth also developed multiple regression equations to predict word length, minimal punctuation unit length, and sentence length.

The Mugford Readability Chart. Mugford (1970) developed a predictive method for readability in the form of a graphic solution much like Fry's. His Readability Chart was intended for 5.5 to 15-year reading ages but has been extended to cover material for adults. It uses the common variables of word-length in syllables and sentence-length in words, but it also takes repetition into account.

Harris-Jacobson Readability Formulas. Harris and Jacobson (1974) developed three readability formulas for primary grade materials. They were based on 481 samples from 56 books. These comprised all of the primary grade pupil books in six widely used series of basal readers and totalled 97,868 words.

Formal procedures have been used to assess the readability of adult magazine materials (Dulin, 1968), vocational materials (Williams, 1979), industrial education textbooks (Clark, 1978), business communication textbooks (Razek & Cone, 1981), economics textbooks (McConnell, 1982), and occupational educational textbooks (New York State Educational Department, 1982). The United States Army evaluated the usefulness of readability formulas for identifying material that would be comprehended by readers at a given reading skill level (Kern, 1980). Extension of this approach was outlined by Marshall (1979) who provided guidelines for analyzing and evaluating textbooks based on comprehensibility rather than readability. Use of readability formulas for writing adult materials to a desired reading level, matching readability levels of material to the reader's skill level, and use of these concepts in adult education was documented by Abram (1981). However, the review of research literature did not reveal application of these procedures to standardized testing instruments. However, text items must also be written at appropriate reading difficulty levels if test scores are expected to yield valid results.

Systematic procedures should be considered for implementation in evaluating written materials. The Dale-Chall (1948), Harris-Jacobson (Harris & Sipay, 1975), Fry (1972), Fog (Klare, 1963), and Flesch (1951) are the most suitable and have the greatest potential for use with an adult audience. These formulas represented the two major approaches to the calculation of readability: estimating the number of unfamiliar words versus the number of syllables per word in addition to sentence length. Although these formulas have been used with adult audiences, only minimal attention has been given to the reading difficulties and characteristics of the ABE learner when responding to standardized tests.

BACK / NEXT PAGE