A Second Look at the Validity of Readability Formulas.
George R. Klare · Journal of Reading Behavior · 1976
In this paper major factors interacting with readability measures in validity studies are identified and described. The basis for the analysis presented are thirty-six experimental studies of the effect of readability variables upon reader comprehension and/or retention. Emphasis is placed on the interacting nature of the factors involved. When first asked to present this paper, I suggested a title different from the one finally selected. My suggestion, I felt, might be catchier, and would certainly result in a briefer address. Title and talk were to be Does Readability Make a Difference? Sometimes. The paper turned out to be a bit longer than that, but the answer, you will see, was an appropriate summary. And the title finally selected, A Second Look at the Validity of Readability Formulas, is also appropriate. Here is why, in a nutshell. Table 1 is a reminder that there are two chief ways of looking at the validity of readability formulas. Most research workers have looked at the first way, which involves the prediction of readable writing. This refers to a formula's ability to assign accurate scores to a large number of different passages of known difficulty. I want to emphasize the second way of looking at validity, which involves the production of readable writing. This refers to modifying a passage into more and/or less readable versions according to formula score, then checking to see if they yield a measurable difference in reader performance or behavior. Before turning to the second way of looking at validity, however, let me say something briefly about the first. An invited address presented at the annual meeting of the National Reading Conference, St. Petersburg, Florida, December 5, 1975. Reprints may be requested from Dr. Klare, Porter Hall, Department of Psychology, Ohio University, Athens, Ohio, 45701.