Automated versus Human Scoring: A Case Study in an EFL-Context.

Autor/in	Huang, Shih-Jen
Titel	Automated versus Human Scoring: A Case Study in an EFL-Context.
Quelle	In: Electronic journal of foreign language teaching, 11 (2014) Supplement 1, S. 149-164 PDF als Volltext kostenfreie Datei Link als defekt meldenVerfügbarkeit
Beigaben	Anhang; Tabellen 5
Sprache	englisch
Dokumenttyp	online; Zeitschriftenaufsatz
ISSN	0219-9874
Schlagwörter	Empirische Forschung; Computer; Essay; Vergleichende Analyse; Fremdsprachenunterricht; Englischunterricht
Abstract	One major development of computer technology involving English writing is automated essay scoring (AES). Previous research has investigated different aspects of AES in writing assessment, such as human and auto-mated scoring differences (Bridgeman, Trapani, & Yigal, 2012), and students' essay structure identification (Burstein & Marcus, 2003). This study addresses two research questions. First, how does automated scoring differ from human scoring in EFL writing? Second, what are EFL learners' perceptions of AES and its effec-tiveness? The instruments involved in this study include an AES system developed by Educational Testing Service (ETS), Criterion, and a survey. The findings of the study suggest that the AES and human scoring are weakly correlated. Besides, the study also finds that an AES system such as Criterion is subject to deliberate human manipulation and can suffer from insufficient explanatory power of computer-generated feedback. The pedagogical implications and limitations of the study are also discussed. (Verlag).
Erfasst von	Informationszentrum für Fremdsprachenforschung, Marburg
Update	2022/2