Analysis of the Rater Effects in the Rating of Diagnostic Trees Prepared by Teacher Candidates By the Many-Facet Rasch Model

Abstract

In the study, it was aimed to investigate the leniency/severity, bias and halo effect of the raters which were used in the rating of the diagnostic tree prepared by the teacher candidates by the many-facet Rasch model. The research study group constitutes 24 teacher candidates who are taking measurement and evaluation courses from the students of school teaching faculty of the education department in a state university. Candidates teachers in the study team have formed two groups between themselves. We have developed a diagnostic tree for each group of fields. Diagnostic trees prepared by teacher candidates were scored by a faculty member in the direction of the same criteria and 12 peers selected from each group. The effects of rating were determined and the data were analyzed with the many-facet Rasch model. As a result of the study, it has been noted that some raters rated severely on some criteria and groups than expected, and some raters showed halo effects on an individual-level.

Keywords: Diagnostic tree, Many-facet Rasch model, Leniency/severity, Bias, Halo effect