Logical versus Empirical Readability and Human Interest Scores: Refinement and Update

Abstract

Two studies, one using 27 and the other 25 general psychology textbooks published between 1982 and 1989, compared several readability and human interest scores for topics presented in the appropriate context. In both cases, Flesch's (logical) scores were significantly correlated (median Pearson r of .49) with students' (empirical) ratings but insufficiently for justifying interchanging them. Another finding was that neither the students' ratings nor Flesch's scores based on the selections yielded significant correlations (median Pearson r of .21) with the corresponding Flesch scores for the books as a whole. An adequate profile of a book may require a number of readability and/or interest scores instead of single global measures.