RESPONSE LATENCY EFFECTS ON CLASSICAL AND ITEM RESPONSE THEORY PARAMETERS USING DIFFERENT SCORING PROCEDURES
Faisal A. Abdelfattah · OhioLink ETD Center (Ohio Library and Information Network) · 2007
Test performance may have little or no personal consequences for examinees in numerous testing situations.In such situations, test administrators are likely to view the effort levels of examinees as a matter for concern.Response latency or item response time was used in an attempt to detect the occurrence of a type of guessing that is expected to occur if the examinee does not put forth adequate effort.That is, responding to an item without adequately reading it and engaging in solution-oriented behavior.The study employed response latency as an indicator of examinees rapid-guessing behavior based on response latencies lower than a particular threshold for each item.The identified responses were rescored differently using the Omitted, the Not-presented, or the Zero scoring procedures.The objective was to evaluate the use of different scoring procedures on Classical Test Theory (CTT) and Item Response Theory (IRT) parameter estimates for dichotomously scored items that were obtained from a sample of 586 ninth grade students from five high schools in Jordan who took a computer-administration of a mathematics test.I would like to express my gratitude to my advisor, Prof. George Johanson, for his unlimited support throughout my study.Without his patience, encouragement, and guidance I could not complete this work.George was always there to meet and talk about many problems I faced.