Composition Correctness Scores

Abstract

A simple teacher-scored method can be used to determine the proportion of correct usage in freshman ESL compositions. The method correlated highly with impressionistic rankings of overall writing quality and discriminated among four very narrow-range levels of proficiency: intensive English, minimally accepted freshman ESL, lower-level ESL, and a parallel lower-level freshman English class (remedial) for native speakers. Within this narrow range of skills, the scoring method provided reliable discrimination for purposes of class assignment and measurement of in-class progress. Six teachers of varying experience in ESL and freshman composition achieved high correlations both as a group and individually among numerical and impressionistic rankings of 20 selected essays, all on expository themes. All scoring was individually determined without group consultation or group grading practice. The two career freshman composition teachers showed slightly lower correlations among all data than the four career ESL teachers, indicating that discussions of criteria might have improved consistency between these two professional groups. Correctness scores are easy to compute and appear to be psychologically convincing to those teachers who prefer an exact numerical score to impressionistic rating as a measure of 1) the seriousness of a given usage error, 2) progress in usage improvement, and 3) readiness to pass into a higher level class. One year's trial and theoretical implications for composition research are discussed.