Please use this identifier to cite or link to this item: https://repository.isls.org/handle/1/8746
Full metadata record
DC FieldValueLanguage
dc.contributor.authorCondor, Aubrey
dc.date.accessioned2023-04-12T20:55:57Z-
dc.date.available2023-04-12T16:15:13Z
dc.date.available2023-04-12T20:55:57Z-
dc.date.issued2022
dc.identifier.citationCondor, A. (2022). Empirically-driven multiple choice response options. In Chinn, C., Tan, E., Chan, C., & Kali, Y. (Eds.), Proceedings of the 16th International Conference of the Learning Sciences - ICLS 2022 (pp. 2122-2123). International Society of the Learning Sciences.en
dc.identifier.urihttps://dx.doi.org/10.22318/icls2022.2122
dc.identifier.urihttps://repository.isls.org//handle/1/8746-
dc.descriptionPosteren
dc.description.abstractWe introduce an empirical method for generating multiple choice (MC) response options for educational assessment. Developing quality response options for MC questions is labor intensive, especially when parallel items are necessary for repeated assessments or multiple test forms. We propose a data-driven, iterative approach for creating an item bank of MC questions by first collecting a sample of open-ended student responses that are subsequently rated by subject-matter experts. Contextual sentence embeddings of the observed student responses as well as the expert ratings are used in conjunction with hierarchical clustering to create MC options that span the desired ground truth rating levels specified by the item creator. We evaluate the quality of the automatically generated response options qualitatively, based on existing research relating to best practices for item writing.en
dc.publisherInternational Society of the Learning Sciences
dc.relation.ispartofurn:ISBN:978-1-7373306-5-3
dc.relation.ispartofProceedings of the 16th International Conference of the Learning Sciences - ICLS 2022, pp. 2122-2123
dc.subjectLearning Sciencesen
dc.titleEmpirically-Driven Multiple Choice Response Optionsen
dc.typeconferencePaper
dc.identifier.doi10.22318/icls2022.2122
Appears in Collections:ISLS Annual Meeting 2022

Files in This Item:
File SizeFormat 
ICLS2022_2122-2123.pdf498.42 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.