Search
Program Calendar
Browse By Day
Browse By Time
Browse By Person
Browse By Room
Browse By Unit
Browse By Session Type
Search Tips
What to do in Chicago
Personal Schedule
Sign In
X (Twitter)
Session Type: Symposium
This session focuses on psychometric modeling of ratings, with particular attention to rater effects. The first paper examines the impact of extensive missing data on the statistical error rates associated with several latent trait model statistics employed for detecting rater effects. The second paper also looks at the issue of missing data in operational rating projects by considering whether full-information maximum likelihood estimation allows for the recovery of interrater reliability coefficients in sparse data matrices. The third paper presents a tri-factor latent trait model which is suitable for capturing variability in rating data contributed by examinees, raters, and items. The fourth paper explores the application of the hyperbolic cosine unfolding model to the detection of rater effects.
The Detection of Severity and Centrality in Raters Under Various Levels of Double Scoring - Rose Stafford, The University of Texas - Austin; Edward W. Wolfe, Pearson; Jodi M. Casabianca, The University of Texas - Austin; Tian Song, Pearson
Estimating Interrater Reliability Using Latent Variable Modeling and Incomplete Data - Grant B. Morgan, Baylor University; Robert L. Johnson, University of South Carolina; Kari Hodge, NACE International Institute
Trifactor Model for the Multiple Ratings Data - Hyo Jeong Shin, University of California - Berkeley; Mark R. Wilson, University of California - Berkeley
Evaluating Rater Accuracy With a Hyperbolic Cosine Unfolding Model - Jue Wang, University of Science and Technology of China; Edward W. Wolfe, Pearson; George Engelhard, University of Georgia