       Document 0753
 DOCN  M9620753
 TI    How reliable is computerized assessment of readability?
 DT    9602
 AU    Mailloux SL; Johnson ME; Fisher DG; Pettibone TJ; Kenai Care Center,
       Alaska, USA.
 SO    Comput Nurs. 1995 Sep-Oct;13(5):221-5. Unique Identifier : AIDSLINE
       MED/96005263
 AB    To assess the consistency and comparability of readability software
       programs, four software programs (Corporate Voice, Grammatix IV,
       Microsoft Word for Windows, and RightWriter) were compared. Standard
       materials included 28 pieces of printed educational materials on human
       immunodeficiency virus/acquired immunodeficiency syndrome distributed
       nationally and the Gettysburg Address. Statistical analyses for the
       educational materials revealed that each of the three formulas assessed
       (Flesch-Kincaid, Flesch Reading Ease, and Gunning Fog Index) provided
       significantly different grade equivalent scores and that the Microsoft
       Word program provided significantly lower grade levels and was more
       inconsistent in the scores provided. For the Gettysburg Address,
       considerable variation was revealed among formulas, with the discrepancy
       being up to two grade levels. When averaging across formulas, there was
       a variation of 1.3 grade levels between the four software programs.
       Given the variation between formulas and programs, implications for
       decisions based on results of these software programs are provided.
 DE    Analysis of Variance  *Artificial Intelligence  Comparative Study
       Evaluation Studies  *Health Education  *Reading  Support, U.S. Gov't,
       P.H.S.  JOURNAL ARTICLE

       SOURCE: National Library of Medicine.  NOTICE: This material may be
       protected by Copyright Law (Title 17, U.S.Code).

