READABILITY RESEARCH

The analysis or measurement of written material in terms of level of difficulty or other characteristics such as human interest The major object of readability research is to simplify the reader’s task. This can be accomplished in part by improving the layout and legibility of the printed page, but the major emphasis of psychologists and reading specialists has been on the content and form of the language used. The choice of words is very important, since vocabulary is one of the major factors in reading ease. Copywriters and others concerned with communication sometimes consult the Teachers Wordbook of 30,000 Words compiled by Thorndike and Lorge. It contains an alphabetical list of the more common English words with their frequency of use; the list is based on exhaustive word counts of different types of material ranging from cookbooks to classics, from postal regulations to newspapers, textbooks, and magazines. Needless to say, a word with a frequency of 100 times per million will be more likely to be understood than a word with a one in a million frequency. The wordbook also contains lists of rare words with even smaller frequencies as well as lists of the 511,000 most common words.Another method used to determine readability is to devise and apply a “readability formula” to measure reading ease. Various formulas have been suggested based on characteristics which have proved to be most closely associated with reading ease; for example, word length, number of abstract words, and sentence structure. The formula in widest use today was developed by Rudolf Flesch (1948, 1949). It takes into account average word length in syllables (wl) and average sentence length in words (si). These elements are combined to give a reading ease (RE) index based on analysis of 100-word passages: RE=206.835—846 wl—1.015 si. An RE score of 0 to 30 is rated very difficult since there are 192 or more syllables per 100 words, and an average sentence length of 29 words or more. The typical magazine presenting such material is described as scientific, and the reader usually has to be a college graduate. An RE score of 70 to 80 is rated fairly easy, since the material contains only 139 syllables per 100 words and fourteen words per sentence. Such material is found in slick-fiction magazines which can be easily read by a seventh grader.Flesch has also devised a human interest score (HI), based on the percentage of personal words (pw) in the passages analyzed—that is, proper names, personal nouns, and personal sentences (ps), which include spoken sentences, questions, and exclamations. The formula is HI=3.635 pw+.314 ps. In applying this formula anything which produces a 60 to 100 HI score is described as dramatic; such material is typically found in fiction magazines. A 0 to 10 HI score, produced by the kind of material in scientific magazines, is rated dull.The Flesch formulas have focused attention on the importance of readability, and have probably contributed to a simplification of style among many journalists, copywriters, and others who write for mass audiences. They have also been utilized in a number of investigations of industrial communications, advertising, and magazine content. Lauer and Paterson (1951) showed that a typical management-union contract had a reading ease score in the very difficult range, requiring a college education. Tiffin and Walsh (1951) analyzed fifty-nine union-management agreements and found that the language was far above the educational level of about 70 per cent of those who were expected to read them.A study of issues of Time and Newsweek made by Trencherd and Crissy (1952) suggests that advertisers are attempting to adapt the level and style of their material to the editorial content. Although they found that the advertising copy was in general easier to read than the editorial copy, a comparison of issues published before and after World War II revealed that the two kinds of copy are coming closer together, with the advertising getting more difficult and the editorial material easier. In another study, Haskins (1960) scored the articles in a single issue of The Saturday Evening Post according to another Flesch formula, the Abstraction Index, and found a high negative correlation between abstraction and a “finishing index”—that is, the more abstract the article, the less likelihood that- the reader would finish it—although those who did finish the more abstract articles were likely to rate them “excellent.” Split-run tests using easy and hard versions of the same articles have reported changes in the volume of readers for the simpler versions ranging all the way from a slight loss to a 66 per cent gain. The higher gains are somewhat questionable, however, because the more difficult material tended to be stilted and less clearly organized.Readability formulas are not universally accepted. While some specialists regard them as a sure guide to better writing, others claim that the “computer” approach makes writing a mechanical affair and completely ignores the esthetic values implicit in figures of speech and unusual vocabulary. Moreover, scientific and literary material should not be rated on the same scale, especially a scale that overlooks the purpose of the material. Can we really call scientific writing “dull” simply because it lacks human interest? It would seem, therefore, that Flesch’s approach has a limited application. As Anastasi (1964) concludes: “Readability formulas are appropriate when the object is to measure the communication of simple messages. This is the purpose of advertising copy, cookbooks, instruction sheets, training manuals government bulletins, and similar forms of writing designed for mass audiences.”