افاق للعلوم
Volume 9, Numéro 4, Pages 535-548
2024-09-29
Using Principal Components Analysis In Designing Academic Writing Scoring Rubrics
Authors : Mizab Manel .
Abstract
Academic writing assessment is pivotal in university career. Some teachers tend to assign holistic marks to written compositions, ignoring assessment criteria. However, this paper highlights the importance of systematicity in assessment using scoring rubrics that are designed according to the context and population they are directed to. This paper suggests a paragraph scoring rubric after performing exploratory Principal Components Analysis (PCA). A set of 19 evaluation criteria is suggested and piloted on 51 paragraphs written by first-year university students from the Department of English at Tebessa University. After setting oblique rotation, PCA indicates that the variables are correlated (sig. 0.001 in Bartlett’s test), and are extracted into five factors that are correlated as well (r 0.32). The PCA pattern matrix indicates that some variables have cross-loadings, so they should be deleted. Therefore, the initial variables are extracted into four factors, named academic writing style, paragraph structure, lexis, and writing mechanics. These components constitute the holistic scoring rubric, which is recommended in scoring paragraphs, and which is tested to validate its groupings and psychometric properties.
Keywords
analytical rubric ; assessment ; holistic rubric ; PCA ; writing
Les articles similaires
Benyounes Djahida
.
pages 584-602.
Boumaza Halima
.
Msserhi Mahbouba
.
pages 99-118.
Adjout Abdelkader
.
pages 155-180.
أبو عبيد أحمد علي خلف
.
ص 25-57.
