1 An Investigation of The An Investigation of The Response Time for Maths Response Time for Maths Items in A Computer Items in A Computer Adaptive Test Adaptive Test C. Wheadon & Q. He, CEM CENTRE, DURHAM UNIVERSITY, UK Chris Wheadon and Qingping Chris Wheadon and Qingping He He ( ( CEM Centre, Durham University, UK CEM Centre, Durham University, UK ) )
21
Embed
1 An Investigation of The Response Time for Maths Items in A Computer Adaptive Test C. Wheadon & Q. He, CEM CENTRE, DURHAM UNIVERSITY, UK Chris Wheadon.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
An Investigation of The Response An Investigation of The Response Time for Maths Items in A Computer Time for Maths Items in A Computer
Adaptive Test Adaptive Test
C. Wheadon & Q. He, CEM CENTRE, DURHAM UNIVERSITY, UK
Chris Wheadon and Qingping HeChris Wheadon and Qingping He
((CEM Centre, Durham University, UKCEM Centre, Durham University, UK))
2
Background
• The CEM Centre’s Baseline Assessments on Primary and Secondary School Students
• The CEM Baseline Tests for Secondary Schools (for students aged 11-18: three separate projects) Paper based, Efficiency, Cost
• Use of CTT and IRT in Computerised Testing and to Interpret Test Data
• To Develop A Computerised Test to Replace the Existing Paper-based Baseline Tests for the Three Secondary Projects
C. Wheadon & Q. He, CEM CENTRE, DURHAM UNIVERSITY, UK
3
• CABT comprises a calibrated item bank and a question
display system
• The adaptive test contains an adaptive maths test and
an adaptive English vocabulary test
• Testing is delivered through the Web or from the
school’s local area network
CABT-The CEM Computer Adaptive Baseline Test
C. Wheadon & Q. He, CEM CENTRE, DURHAM UNIVERSITY, UK
4
Establishing A Calibrated Item Bank
C. Wheadon & Q. He, CEM CENTRE, DURHAM UNIVERSITY, UK
• Testing of items by school students from different year groups through the administration of a series of tests containing common items
• Calibration of items in each test using the Rasch model (items meeting the Rasch requirements are used)
• Setting of the reference test and equating of different tests using the Rasch model and common items
• Further new item calibration through the embedding of new items in the adaptive test
5
Developing The Question Display System - Conducting Adaptive Testing
C. Wheadon & Q. He, CEM CENTRE, DURHAM UNIVERSITY, UK
• Realisation of adaptivity is achieved through the implementation in the question display system of the Rasch model for ability estimation (MLE method is used) and question selection
• Rules for stopping test: minimum number of question; maximum number of questions; and convergence value
• Variable starting difficulty for the first question for different year groups.
6
C. Wheadon & Q. He, CEM CENTRE, DURHAM UNIVERSITY, UK
Preliminary Results from theAdaptive Maths Test
• There are over 500 items in the maths item bank
• Effort has been made to make the items as curriculum-free as possible
• Items are content-independent to each other
• Item types include MCQ, short free text entry questions and interactive questions
• Items cover a wide range of difficulties in order for the three secondary projects (for three different year groups) to use the same adaptive test as their baseline test