- Browse
Search
On-Site Program Calendar
Browse By Day
Browse By Time
Browse By Person
Browse By Room
Browse By Unit
Browse By Session Type
- Information Menu
Search Tips
Annual Meeting Housing and Travel
- Navigation and Settings Menu
Sign In
Differential Item Functioning Detection Using Data Mining Techniques: Identifying Additional Sources of Unfairness in Educational Measurement
Sat, April 23, 4:15 to 5:45pm PDT (4:15 to 5:45pm PDT), Division Virtual Rooms, Division D - Section 1: Educational Measurement, Psychometrics, and Assessment Virtual Paper Session RoomAbstract
Traditional DIF detection methods and the alternative method Rasch tree are used to examine DIF in a Rasch model with the usage of 16 items of science literacy test from 2018 PISA. DIF analyses are conducted based on students’ gender (Female or Male), origin of country (United States or Taiwan), and index of economic, social and cultural status (ESCS), which is a continuous variable provided by PISA. Traditional methods reach the same conclusions with respect to DIF items for categorical covariates, while inconsistent results are found for ESCS. Rasch tree identifies six DIF sub-samples and offers more insights into the underlying dynamics that cause the unfairness in item responses. Results obtained from different methods are compared and implications are discussed.