Recognizing Issues of Readability Formulas
As the goal of technical writers is to connect and communicate clearly with their audiences, a great deal of scholarship has been dedicated to discovering what criteria actually facilitate reading comprehension for varying audiences. As early as the 1920s, researchers began to develop readability formulas in order to indicate the difficulty of texts for readers. Janice Redish and Jack Selzer provide an overview of these readability formulas, describing them as "mathematical equation[s] that [are] applied to prose texts in an effort to predict how difficult the text will be for a given group of readers. When you apply a readability formula, you get a score (either a number from 0 to 100 or a reading grade level, depending on the formula)" (Redish and Selzer 1985). However, Redish and Selzer, in addition to Alan Bailin and Ann Grafstein, explain that readability formulas provide little scientific evidence for enhancing comprehension in readers. Instead, these researchers provide evidence that suggests that using readability formulas to design clearer, easier to read texts, can actually hinder the audience's comprehension of the texts. The sections below will outline some issues that readability formulas raise and will discuss some solutions for creating more user-centered texts as suggested by Redish, Selzer, Bailin, and Grafstein.
The Issue with Readability Formulas
The main issue with readability formulas is the fact that they are formulas. As Redish and Selzer indicate in their article "The Place of Readability Formulas in Technical Communication, "because only features that can be counted can be included in the [readability] formula[s], many features that experts agree are important were not included in the formulas developed" (Redish and Selzer 1985). Thus, the criteria considered in readability formulas is limited to only two of several important features in a text: word difficultly and sentence difficulty (Redishand Selzer 1985, Bailin and Grafstein 2001). Moreover, the developers of these formulas faultily measure difficulty of both words and sentence based on their length. This proves to be problematic in defining the readability of a text as not all shorter words or shorter sentences are necessarily easier to understand than longer ones. To exemplify this, Redish and Selzer present the short technical term "waive", which would appear on the same word list as "we," "will", and "your" in many readability formulas simply because it contains the same number of syllables (Redish and Selzer 1985). Similarly, Redish and Selzer demonstrate how this notion would carry over to sentence length, by providing the following two passages:
- The defendant is a fifteen-year-old teenager who is accused of shoplifting.
- He is the defendant. He is fifteen years old. He is in his teens. Someone says he stole from the store. (Redish and Selzer 1985).
The first sentence, although longer, enables readers to make connections between the facts of the defendant. However, the second passage is broken down into four very short sentences, which slows the reader's process in comprehending the facts about the defendant. Here, readability formulas fall through in enhancing reader comprehension because it forces the writer to break down ideas that are causally related and requires the reader to rebuild this connection.
Bailin and Grafstein elaborate on the facilitation of comprehension by discussing the significance of writing sentences with coordinated conjunctions and logical connectives. By including words such as "but," "and," and "because" in passages, writers can help their readers recognize a communicative unity between clauses such as "I couldn't answer your e-mail" and "There was a power outage" (Bailin and Grafstein 2001).
Many other problems result from readability formulas simply because they are mechanistic and, today, even computerized. Thus, as Redish and Selzer, state, "They do not interpret context, meaning, grammar, or content." Readability formulas, although they can provide readers with a very general sense of how difficult a text may be, cannot suggest how very specific users (the elderly, or mentally disabled, or even general adult audiences) may experience and comprehend a text. In the next section, we will discuss some solutions for the production of more readable texts as suggested in the works of Redish and Selzer and Bailin and Grafstein.
As word and sentence length do not prove to be suitable criteria for assessing readability of texts, we must turn instead to new methods for both assessing readability and writing with the reader's comprehension in mind. The following two sub-sections first address stylistic solutions for writers and then present alternative methods to testing the readability of tests.