Validating a computerized scoring system for assessing writing and placing students in composition courses

How do scores from writing samples generated by computerized essay scorers compare to those generated by “untrained” human scorers and what combination of scores, if any, is more accurate at placing students in composition courses? This study endeavored to answer this two-part question by evaluating...

Full description

Saved in:
Bibliographic Details
Published in:Assessing writing 2006, Vol.11 (3), p.167-178
Main Author: James, Cindy L.
Format: Article
Language:English
Subjects:
ACCUPLACER
Automated scoring
College Faculty
Computer Assisted Testing
Correlation
Essays
IntelliMetric
Placement
Predictive Validity
Scoring
Student Placement
Validity
Writing (Composition)
Writing assessment
Writing Instruction
Writing Tests
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:How do scores from writing samples generated by computerized essay scorers compare to those generated by “untrained” human scorers and what combination of scores, if any, is more accurate at placing students in composition courses? This study endeavored to answer this two-part question by evaluating the correspondence between writing sample scores generated by the IntelliMetric™ automated scoring system and scores generated by University Preparation English faculty, as well as examining the predictive validity of both the automated and human scores. The results revealed significant correlations between the faculty scores and the IntelliMetric™ scores of the ACCUPLACER™ OnLine WritePlacer Plus test. Moreover, logistic regression models that utilized the IntelliMetric™ scores and average faculty scores were more accurate at placing students (77% overall correct placement rate) than were models incorporating only the average faculty score or the IntelliMetric™ scores.
ISSN:1075-2935
1873-5916
DOI:10.1016/j.asw.2007.01.002