Automated assessment of student learning has become the subject of increasing attention. Students’ textual responses to short answer questions offer a rich source of data for assessment. However, automatically analyzing textual constructed responses poses significant computational challenges, exacerbated by the disfluencies that occur prominently in elementary students’ writing. With robust text analytics, there is the potential to analyze a student’s text responses and accurately predict his or her future success. In this paper, we propose applying soft cardinality, a technique that has shown success grading less disfluent student answers, on a corpus of fourth-grade responses to constructed response questions. Based on decomposition of words into their constituent character sub-strings, soft cardinality’s evaluations of responses written by fourth graders correlates with summative analyses of their content knowledge.
Leeman-Munk, S., Shelton, A., Wiebe, E. N., & Lester, J. C. (2014). Towards Domain-Independent Assessment of Elementary Students’ Science Competency using Soft Cardinality. In Proceedings of The 9th Workshop on Innovative Use of NLP for Building Educational Applications. Baltimore, Maryland.