Bilimsel süreç becerilerin ölçülmesinde bütünsel ve analitik puanlama anahtarlarının güvenirliklerinin karşılaştırılması
Özet
This study aims to compare the interrater reliability levels of scoring based on the holistic and the analytic scoring rubrics which evaluates the scientific process skills. A total of 100 students, including 62 girls and 38 boys has participated the research from the 6-A, 6-B and 6-D classes of the Ankara Province Mamak Hamdi Bulgurlu Primary School. The evaluation of the test that are applied to students has been conducted by two science and technology teachers. A test has been used consist of four open-ended items prepared on the basis of the science questions placed in the TIMSS 2007 ?Science Test. This test aims to assess students? scientific process skills who are in the 6th class. These questions has been prepared for the assessment of the scientific process skills which are mainly construct of hypothesis, controlling and manipulating of variables and, prediction of results, data interpretation and conclusion. Besides, holistic and analytic rubric have been used which are formed in order to score answers of test. Data which has been collected to answer the questions stated in line with the aims of the research, has been analyzed with using the SPSS 12.0 software package program. In this context, the Spearman-Brown was computed in terms of reliability coefficient for interraters and the Pearson Product-Moment Correlation coefficient has been calculated for relationship between science lesson?s mark and rubric score.This study results that the consistency among the scores of the graders with using the analytical scoring rubric is more than the consistency of the scores of the graders with using the holistic scoring rubric. Besides, there exists a significant relationship at mid-level between the Science and Technology lesson?s mark and both rubric scores, the result of which the 40% of the course achievement grades can be explained by these test scores.