Reliability and validity study of the achievement scale of electromechanic systems course


Yilmaz O., Tuncalp K., Arslan A.

2nd World Conference on Educational Sciences (WCES-2010), İstanbul, Turkey, 4 - 08 February 2010, vol.2, pp.2722-2727 identifier

  • Publication Type: Conference Paper / Full Text
  • Volume: 2
  • Doi Number: 10.1016/j.sbspro.2010.03.403
  • City: İstanbul
  • Country: Turkey
  • Page Numbers: pp.2722-2727

Abstract

The purpose of the study of Achievement Scale of Electromechanic Systems (ASES) is to introduce the achievements of the students studying this course, in their learning-teaching processes, and to contribute to evaluation and assessment of the acquisitions in this field. At the beginning of the reliability and validity study of the achievement scale, a question pool consisting of 100 questions was formed, by considering the subjects stated in the course program of the Electromechanic Systems Course as well as the target behaviors. In line with the opinions of the academicians and evaluation and assessment experts active in the field of Electromechanic Systems, some (60) of these questions were selected and used in the study. The study was carried out in two stages. In the first stage, a group consisting of 95 students studying this course in Marmara University Technical Education Faculty was subjected to achievement test and validity analysis. In consequence of this validity analysis, the items (20), validity levels of which were low and item difficulties of which remained out of the scientific criteria, were eliminated and the achievement test was applied again to the same group (95 students). Acceptable state (40) and correlation (r) of the achievement test of these two applications were considered. High-level positive relation was observed between the two tests (r = 0.881). The 20 questions over the significance level of 0.05 in all the item-total, item-remainder and item-distinctiveness relations were eliminated and the number of questions was reduced to 40. Cronbach alpha coefficient was found 0.802, calculated statistically as based on the variance of each question. In the second stage, the valid and reliable achievement test was applied to the same group consisting of 95 students 8 weeks later. High-level positive relation was observed between the two measurements of the valid and reliable achievement tests repeated 8 weeks later (r=0.881, p<0.001). (C) 2010 Elsevier Ltd. All rights reserved.