Readability

Readability is generally defined by English-speaking literacy scholars as the level of knowledge and skill required to make full sense of a given printed text. This view of readability is most evident in formulas such as the Flesch Ease of Reading that was developed in 1948 based on the assumption that the fewer words in a sentence and the more familiar these words are in a given text, the less difficult it is for readers to comprehend this text. Word familiarity is an indirect yet stable indicator of students’ ability to comprehend a word, which in turn has an effect on a reader’s comprehension of a given text. Furthermore, the more simple and brief the sentence structures within a text, the greater the ease for readers to understand the intended messages carried within such structures.

The most currently ubiquitous readability indicator is the Flesch-Kincaid, which is essentially a revised version of the original Flesch formula, producing a grade level as its readability score. Originally developed by the U.S. Navy in 1975 to determine the relative difficulty of their various technical manuals, the Flesch-Kincaid formula has become an integral component of many widely used online reading programs and linguistic analytic tools.

Like other readability algorithms, the Flesch-Kincaid determines the word-level familiarity of a printed text by the average frequency of individual words (i.e., the likelihood that a reader would be exposed to a particular word based on the analysis of a corpus of books read by adults) and the average sentence length. The lower the average likelihood, or frequency value of words presented in the text, the more difficult the text is deemed for readers. Similarly, the longer the average sentence length, the more assumed difficulty in comprehending key points presented in embedded sentence structures. Simply put, the more frequently a word occurs in a language, the greater the likelihood that students will know its meaning. However, high-frequency words tend to denote more general concepts or categories such as man or work, rather than more specialized words like radiologist or employment. Thus, it may be argued that the more frequent a word, the greater the likelihood that while students will know its meaning, this meaning may be less precise than what was intended in the text.

Applications of readability or text analytic software for analysis, research, or text development purposes generally follow more qualitative efforts to achieve textual accuracy, coherence, and [Page 1381]meaningfulness to readers. Even quantitative programs focused on determining textual cohesion can only do so at a lexical level; that is, the extent to which ideas presented in a text support one another can only be determined by a reader. Thus, while readability indicators offer a general idea about the difficulty of printed texts, such metrics should not be the sole guides for text development.