Wednesday, September 1, 2010

Research in English: Validity of Holistically Scored Writing Samples


Purpose of this blog: Reviews of interesting research in English education journals. When possible, I suggest how I might apply the findings to my own classroom.


10-second review: Validity of Holistically scored writing samples depends on purpose. For placement, validity is acceptable. For purposes like preventing students from graduating, cannot be defended.

Title: “Locally Developed Writing Tests and the Validity of Holistic Scoring.” T McKendry. Research in the Teaching of English (May 1992), 149-161.

Quote: “…none of the statistical validity studies, local or large-scale, shows a correlation that is more than moderate….”

Quote: “For purposes like placement, holistically scored writing samples may be accepted as the most practical, least invalid choices. Where possible other measurements, such as previous grades or a second writing sample, can be combined with the holistically scored essay to make placement more accurate.”

Quote: “For tests with more important consequences, such as barrier tests that prevent students from graduating, however, holistically scored writing samples cannot easily be defended on statistical or theoretical grounds.”

Comment: Even for placement, local holistically scored essays should be combined with previous English course grades and other criteria, like objective tests of writing as with the SAT. RayS.

No comments: