Using reading grade level to assess readability of selected plant and soil science textbooks
John G. Graveel, Henry A. Fribourg · Journal of Agronomic Education · 1987
ABSTRACT Commonly used reading grade level (RGL) indices (Coleman, Dale‐Chall, Devereux Automated Reading Index, Flesch Reading Ease Score, Flesch‐Kincaid, FOG, Fry, SMOG) were used to estimate the RGL of selected plant and soil science textbooks used in lower division, upper division, and graduate level college courses. Readability formulas, in addition to many other factors, can be considered when evaluating the suitability of a text for a specific course level. The objectives were to determine whether RGL assessment techniques used for elementary and secondary education texts would discriminate among plant and soil science textbooks, to select the RGL indices suited to quantify the readability of these sources, and to identify the factors contributing to the variability among texts. Ten passages of at least 300 words consisting of complete sentences only were selected at random from each source. Distribution of the RGL data approached normality within each source. Factor analysis yielded a three‐factor (word difficulty, sentence complexity, scientific content) solution, which explained 94% of the total observed variability with the following five variates: Fry graph coordinates, Flesch score, Coleman index, and Dale‐Chall index with and without scientific vocabulary included in the Dale list of familiar words. This study is unique in its inclusion of scientific content to characterize the readability of science textbooks. Discriminant analysis, using the same five variables to generate the discriminant functions, showed that the three student levels (lower division, upper division, graduate) constituted three distinct populations. All the indices can be computed easily with commercially available microcomputer software. However, the selection of the terms to be added to the Dale list may be time‐consuming, depending on the scientific and specialized content of the sample texts. After obtaining the RGL data, the determination of textbook suitability for a particular student population can be pursued through the use of standard statistical analyses.