Comparison of Classification Judgments Scaling Methods


Creative Commons License

Buyukkidik S.

CUKUROVA UNIVERSITY FACULTY OF EDUCATION JOURNAL, cilt.47, sa.1, ss.260-284, 2018 (ESCI İndekslerine Giren Dergi) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 47 Konu: 1
  • Basım Tarihi: 2018
  • Doi Numarası: 10.14812/cuefd.288600
  • Dergi Adı: CUKUROVA UNIVERSITY FACULTY OF EDUCATION JOURNAL
  • Sayfa Sayıları: ss.260-284

Özet

In this study, 264 performance tasks 6, 7 and 8th grade secondary school students answered was examined by the scaling method with classification criteria to determine whether classification judgements differ or not obtained from the scores of 0-15 by four raters with analytical and holistic rubrics ("0-3" beginning level, "4-6" developable, "7-9" apprentices, "10-12" headworker, "13-15" master). The research is based on descriptive research with the dimension of not aiming generalization and revealing the present situation. In the study, data were collected with two performance tasks, including evidence of validity and reliability and rubrics related to these tasks. As a result of the research, it is seen that scale values and orders of classification judgements of four raters for 264 performances in terms of problem solving ability can be partially changed according to rubric type and the method used. While the scaling method does not cause any difference in the order of scale values in the case of using holistic rubric; it is seen that the rank order of the scale obtained from the B full data matrix and the scale sequences obtained from the B rank numerical solution and D state full matrices are different in the case of using analytical rubric.