文章基本信息

标题：Compiling a Test: How to Solve Calibration Issues
本地全文：下载
作者：Anaje Armendariz ; Anaje Armendariz ; Tomás A. Pérez 等
期刊名称：Procedia - Social and Behavioral Sciences
印刷版ISSN：1877-0428
出版年度：2012
卷号：46
页码：2253-2257
DOI：10.1016/j.sbspro.2012.05.464
语种：English
出版社：Elsevier
摘要：AbstractWhen compiling a test, there are several questions that have to be solved before proceeding to include the items. For instance: do I have to evaluate a minimum required knowledge (i.e. pass or fail) or do I just have to classify the examinees? How much accuracy shall I need? Which set of items fits best my needs? How difficult should these items be? How do I ensure that all those items are correct? Do they assess only the contents I want them to? How can I verify that? These issues are well documented in Psychometrics and the answers to the questions posed are usually based on statistic procedures that have to be followed during the so-called item calibration process. CALLIE is a revolutionary system in the field of item calibration. Its main objective is to support the calibration process and therefore let teachers, pedagogues and other content developers without specific background on psychometrics easily calibrate their own item banks. CALLIE includes a variety of workflow processes to solve some of the questions previously posed. These processes are integrated into a bigger workflow process that contains the milestones and decisions of a complete calibration. This paper describes the system and its features.
关键词：calibration process workflow;computer adaptive testing;decision taking supporting tool;item response theory