Comparative Quality Analysis of GPT-Based Multiple Choice Question Generation

[en] Assessment is an essential part of education, both for teachers who assess their students as well as learners who auto-evaluate themselves. A popular type of assessment questions are multiple-choice questions (MCQ), as they can be automatically graded and can cover a wide range of learning items. However, the creation of high quality MCQ items is nontrivial. With the advent of Generative Pre-trained Transformer (GPT), considerable effort has been recently made regarding Automatic Question Generation (AQG). While metrics have been applied to evaluate the linguistic quality, an evaluation of generated questions according to the best practices for MCQ creation has been missing so far. In this paper, we propose an analysis of the quality of automatically generated MCQs from 3 different GPT-based services. After producing 150 MCQs in the domain of computer science, we analyse them according to common multiple-choice item writing guidelines and annotate them with identified docimological issues. The dataset of annotated MCQs is available in Moodle XML format. We discuss the different flaws and propose solutions for AQG service developers.

Disciplines :

Computer science

Author, co-author :

GREVISSE, Christian ; University of Luxembourg

External co-authors :

Language :

English

Title :

Comparative Quality Analysis of GPT-Based Multiple Choice Question Generation

Publication date :

24 October 2023

Event name :

6th International Conference on Applied Informatics

Event organizer :

ITIUD

Event date :

26 to 28 October 2023

Audience :

International

Main work title :

Communications in Computer and Information Science

Publisher :

Springer Nature Switzerland

ISBN/EAN :

978-3-03-146813-1
978-3-03-146812-4

Pages :

435-447

Peer reviewed :

Peer reviewed

Additional URL :

https://link.springer.com/content/pdf/10.1007/978-3-031-46813-1_29

Available on ORBilu :

since 24 October 2023

Statistics

Number of views

99 (17 by Unilu)

Number of downloads

10 (4 by Unilu)

More statistics

Scopus citations^®

Scopus citations^®
without self-citations

OpenCitations

Bibliography

Bandiera, G., Sherbino, J., Frank, J.R.: The CanMEDS assessment tools handbook: an introductory guide to assessment methods for the CanMEDS competencies. Royal College of Physicians and Surgeons of Canada (2006)
Bertrand, C., et al.: Choisir un outil d’évaluation. In: Pelaccia, T. (ed.) Comment (mieux) former et évaluer les étudiants en médecine et en sciences de la santé?, pp. 357–370. De Boeck Supérieur (2016)
Bloom, B.S.: Taxonomy of Educational Objectives: The Classification of Educational Goals. Allyn and Bacon, Boston (1956)
Bongir, A., Attar, V., Janardhanan, R.: Automated quiz generator. In: Thampi, S.M., Mitra, S., Mukhopadhyay, J., Li, K.-C., James, A.P., Berretti, S. (eds.) ISTA 2017. AISC, vol. 683, pp. 174–188. Springer, Cham (2018). https://doi.org/10. 1007/978-3-319-68385-0 15
Cortés, J.A., Vega, J.A., Schotborg, D.C., Caicedo, J.C.: Education platform with dynamic questions using cloud computing services. In: Solano, A., Ordoñez, H. (eds.) CCC 2017. CCIS, vol. 735, pp. 387–400. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-66562-7 28
DiBattista, D., Sinnige-Egger, J.A., Fortuna, G.: The “none of the above” option in multiple-choice testing: an experimental study. J. Exp. Educ. 82(2), 168–183 (2014). https://doi.org/10.1080/00220973.2013.795127
Dijkstra, R., Genç, Z., Kayal, S., Kamps, J.: Reading comprehension quiz generation using generative pre-trained transformers. In: Sosnovsky, S.A., Brusilovsky, P., Lan, A.S. (eds.) Proceedings of the Fourth International Workshop on Intelligent Textbooks 2022 Co-Located with 23d International Conference on Artificial Intelligence in Education (AIED 2022), Durham, UK, 27 July 2022. CEUR Workshop Proceedings, vol. 3192, pp. 4–17. CEUR-WS.org (2022). https://ceur-ws.org/Vol-3192/itb22 p1 full5439.pdf
Gabajiwala, E., Mehta, P., Singh, R., Koshy, R.: Quiz maker: automatic quiz generation from text using NLP. In: Singh, P.K., Wierzchoń, S.T., Chhabra, J.K., Tan-war, S. (eds.) Futuristic Trends in Networks and Computing Technologies. LNEE, vol. 936, pp. 523–533. Springer, Singapore (2022). https://doi.org/10.1007/978-981-19-5037-7 37
Gilal, A.R., Waqas, A., Talpur, B.A., Abro, R.A., Jaafar, J., Amur, Z.H.: Question guru: an automated multiple-choice question generation system. In: Al-Sharafi, M.A., Al-Emran, M., Al-Kabi, M.N., Shaalan, K. (eds.) ICETIS 2022. LNNS, vol. 573, pp. 501–514. Springer, Cham (2023). https://doi.org/10.1007/978-3-031-20429-6 46
Goyal, R., Kumar, P., Singh, V.P.: Automated question and answer generation from texts using text-to-text transformers. Arab. J. Sci. Eng. (2023). https://doi. org/10.1007/s13369-023-07840-7
Haladyna, T.M., Downing, S.M., Rodriguez, M.C.: A review of multiple-choice item-writing guidelines for classroom assessment. Appl. Measur. Educ. 15(3), 309– 333 (2002). https://doi.org/10.1207/S15324818AME1503 5
Kumar, A.P., Nayak, A., Manjula Shenoy, K., Chaitanya, Ghosh, K.: A novel framework for the generation of multiple choice question stems using semantic and machine-learning techniques. Int. J. Artif. Intell. Educ. (2023). https://doi.org/10. 1007/s40593-023-00333-6
Kumar, S., Chauhan, A., Pavan Kumar, C.: Learning enhancement using question-answer generation for e-book using contrastive fine-tuned T5. In: Roy, P.P., Agarwal, A., Li, T., Krishna Reddy, P., Uday Kiran, R. (eds.) BDA 2022. LNCS, vol. 13773, pp. 68–87. Springer, Cham (2022). https://doi.org/10.1007/978-3-031-24094-2 5
Manrique, R., Grévisse, C., Mariño, O., Rothkugel, S.: Knowledge graph-based core concept identification in learning resources. In: Ichise, R., Lecue, F., Kawamura, T., Zhao, D., Muggleton, S., Kozaki, K. (eds.) JIST 2018. LNCS, vol. 11341, pp. 36–51. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-04284-4 3
MoodleDocs: Calculated question type. https://docs.moodle.org/402/en/Calculated question type. Accessed 10 July 2023
Mulla, N., Gharpure, P.: Automatic question generation: a review of methodologies, datasets, evaluation metrics, and applications. Prog. Artif. Intell. 12(1), 1–32 (2023). https://doi.org/10.1007/s13748-023-00295-9
Shank, P.: Write Better Multiple-Choice Questions to Assess Learning: Measure What Matters—Evidence-Informed Tactics for Multiple-Choice Questions. Learning Peaks LLC (2021)
Srihari, C., Sunagar, S., Kamat, R.K., Raghavendra, K.S., Meleet, M.: Question and answer generation from text using transformers. In: Thampi, S.M., Mukhopadhyay, J., Paprzycki, M., Li, K.C. (eds.) ISI 2022. SIST, vol. 333, pp. 201–210. Springer, Singapore (2023). https://doi.org/10.1007/978-981-19-8094-7 15
Vachev, K., Hardalov, M., Karadzhov, G., Georgiev, G., Koychev, I., Nakov, P.: Leaf: multiple-choice question generation. In: Hagen, M., et al. (eds.) ECIR 2022. LNCS, vol. 13186, pp. 321–328. Springer, Cham (2022). https://doi.org/10.1007/978-3-030-99739-7 41