Person fit based on statistical process control in an adaptive testing environment

Edith van Krimpen-Stoop, R.R. Meijer

Research output: Book/ReportReportProfessional

19 Downloads (Pure)

Abstract

Person-fit research in the context of paper-and-pencil tests is reviewed, and some specific problems regarding person fit in the context of computerized adaptive testing (CAT) are discussed. Some new methods are proposed to investigate person fit in a CAT environment. These statistics are based on Statistical Process Control (SPC) theory. A technique from SPC that is effective in detecting small shifts in the mean of the variable being measured is the cumulative sum (CUSUM) procedure. How CUSUM is applied to CATs is outlined, and eight statistics are proposed to investigate the sum of consecutive negative or positive residuals. Two simulation studies evaluated the use of these eight statistics. Results show that the detection rates of these statistics are dependent on the type of aberrance simulated. Conditions under which the statistics can be used or should not be used are discussed.
Original languageEnglish
Place of PublicationEnschede
PublisherUniversiteit Twente TO/OMD
Number of pages23
Publication statusPublished - 1998

Publication series

NameOMD Research Report
PublisherUniversity of Twente, Faculty of Educational Science and Technology
No.98-13

Keywords

  • Simulation
  • Computer Assisted Testing
  • METIS-136531
  • Goodness of Fit
  • IR-103771
  • Adaptive Testing

Fingerprint Dive into the research topics of 'Person fit based on statistical process control in an adaptive testing environment'. Together they form a unique fingerprint.

Cite this