Understanding Average Score Differences Between e-rater and Humans for Demographic-Based Groups in the GRE General Test AES GRE
- Author(s):
- Ramineni, Chaitanya; Williamson, David M.; Weng, Vincent Z.
- Publication Year:
- 2014
- Source:
- Wendler, Cathy; Bridgeman, Brent (eds.) with assistance from Chelsea Ezzo. The Research Foundation for the GRE revised General Test: A Compendium of Studies. Princeton, NJ: Educational Testing Service, 2014, p4.9.1-4.9.5
- Document Type:
- Chapter
- Page Count:
- 5
- Subject/Key Words:
- Graduate Record Examination (GRE), General Test (GRE), e-rater, Revised GRE, Human Raters, Automated Essay Scoring (AES)
Abstract
Examined possible root causes for those discrepancies seen in Chapter 4.8 (of "A Comprehensive Meta-Analysis of the Predictive Validity of the GRE: Implications for Graduate Student Selection and Performance") between scores generated by human raters and those generated by e-rater across various subgroups. The research suggested that e-rater is not severe enough on grammatical language errors (compared to humans), tends to overvalue long essays, and occasionally undervalues content.