Faculty Publications


ERIC Document - ED021878 found in the ERIC Database

Document Type



College Freshmen; Curriculum Research; Educational Experiments; Educational Testing; English; English Curriculum; Evaluation Methods; Grading; Group Testing; Higher Education; Statistical Data; Writing (Composition); Writing Skills


To supplement an experiment at the University of Northern Iowa on the effectiveness of college-level instruction in freshman composition; the University of Iowa conducted its own experiment on the evaluation of freshman composition. This experiment was designed to determine (1) which method of rating papers is more reliable--the "general impression" method of the Educational Testing Service or the "analytical" method used at the University of Iowa; (2) if a more significant measure of writing quality is obtained when the better of two papers; rather than merely one; is used as a criterion; and (3) if the writing performance of students completing the University's rhetoric course improved more than that of students excused from the course. The procedures by which theme raters and experimental and control groups of students were selected are described in detail; as are the methods of rating themes. Chapters discussing findings; conclusions; and recommendations indicate that the results of the experiment were inconclusive because the rating of the themes was not reliable. Included in the report are extensive tables of data and appendixes listing information about and instructions to participants in the experiment. (BN)


Department of Languages, Speech, and Literature

Original Publication Date


Object Description

1 PDF file (77 pages)


UNI ScholarWorks, Rod Library, University of Northern Iowa


©1968 Richard Braddock and Charles R. Statler



File Format