期刊名称：Journal of Educational, Cultural and Psychological Studies (ECPS Journal)
印刷版ISSN：2037-7924
出版年度：2014
卷号：1
期号：9
页码：147-165
DOI：10.7358/ecps-2014-009-luci
语种：Italian
出版社：LED Edizioni Universitarie
摘要：Which Quantity and Quality for Understanding Educational Experiences This paper illustrates the methodology adopted to design and calibrate a test for measuring the student’s linguistic abilities at the end of the third year of vocational training courses (10th grade). The work is part of a major ISFOL study also including tests on math and strategic abilities. The test was calibrated on a judgement sample of 1046 students. It measures listening, reading comprehension, mechanics and written composition through several tasks. This paper presents methodologies of two item analysis procedures the CTT («Classical Test Theory») and IRT («Item Response Theory»), and the procedures for designing derivate variables from different tasks. In particular, it presents the advantages of using the Rasch model for a better measure of the latent trait of ability and for calibrating the test items. The paper rounds off by illustrating the main results of the multivariate analysis of the study. These results should obviously be interpreted in the light of the characteristics of the target population, its social and cultural background, which is weak in this case, and the collection of its school failures.
其他摘要：Which Quantity and Quality for Understanding Educational Experiences This paper illustrates the methodology adopted to design and calibrate a test for measuring the student’s linguistic abilities at the end of the third year of vocational training courses (10th grade). The work is part of a major ISFOL study also including tests on math and strategic abilities. The test was calibrated on a judgement sample of 1046 students. It measures listening, reading comprehension, mechanics and written composition through several tasks. This paper presents methodologies of two item analysis procedures the CTT («Classical Test Theory») and IRT («Item Response Theory»), and the procedures for designing derivate variables from different tasks. In particular, it presents the advantages of using the Rasch model for a better measure of the latent trait of ability and for calibrating the test items. The paper rounds off by illustrating the main results of the multivariate analysis of the study. These results should obviously be interpreted in the light of the characteristics of the target population, its social and cultural background, which is weak in this case, and the collection of its school failures.
关键词：Evaluation, Exams, Measurement, Quality of measures, Text, Esami, Misurazione, Qualità delle misure, Test, Valutazione
