HKIER

Abstract

Of increasing interest to the educational assessment researchers is the role of educational assessment practices on student achievement-related outcomes (Black & Wiliam, 1998). By their inherent nature, the data collected in this line of research are hierarchically structured in that students are nested within classes. As might be expected, not considering the nested nature of the data in the educational assessment research may lead to invalid inferences about the relationship between educational assessment and student motivation and achievement. As a means of drawing valid inferences from hierarchically structured data, this paper highlights the utility and applicability of hierarchical linear modeling techniques in the educational assessment research. These techniques not only facilitate a decomposition of the relationship between the variables into separate student-level and class-level components, but also recognize the dependence among the outcomes of students within the same class (Raudenbush & Bryk, 2002). This dependence may arise as a result of shared students’ experiences with regard to the teacher’s assessment practices. This paper discusses the necessity for using these techniques in the educational assessment research in order to validate inferences and further research agenda in this area.

Keywords: Hierarchical Linear Models; educational assessment research; data analysis

摘要

N.A.