Response Times for Mathematics Items in Computer Adaptive Testing in Secondary Schools

Response Times for Mathematics Items in Computer Adaptive Testing in Secondary Schools

[featured_image]
  • Version
  • Download 87
  • File Size 294.43 KB
  • File Count 1
  • Create Date August 2, 2018
  • Last Updated August 2, 2018

Response Times for Mathematics Items in Computer Adaptive Testing in Secondary Schools

An advantage of computer based testing over conventional paper and pen based testing is that response time to items from those taking the test can be accurately recorded for subsequent analysis. However, when used for testing mathematics, there is a disadvantage in that only the test taker’,s answer is recorded with no information as to how he/she obtained it. In this paper we investigate whether a tendency to guess at answers, rather than reasoning them out, can be discerned from the data. Using a Rasch model to measure difficulty of items and ability of test takers, with a bank of about 500 items in tests taken by about 80000 secondary age pupils (11-15) in England, we conclude that if a certain gap between ability and difficulty is exceeded then pupils are likely to guess. We investigate differences in gender and age of pupil, and type of item. We discuss choice of items in light of the intended purpose of a test and the practicalities of managing such tests in a school environment.

Attached Files

FileAction
paper_1162d348b.pdfDownload