Automated assessment of parametric programming in a large-scale course
Francisco de Assis Zampirolli, Cristiane M. Sato, Heitor Rodrigues Savegnago, Valério Ramos Batista, Guiou Kobayashi · 2021 XVI Latin American Conference on Learning Technologies (LACLO) · 2021
The main contribution of this paper is in the method of creating and correcting parameterized programming exercises. This method was validated by an experiment carried out in a course on Computational Bases of Science that was given during the 2020 Coronavirus Pandemic. Lists of individualized exercises were emailed to the students every week through an adaptation of the open system MCTest, and corrections were made automatically by an adaptation implemented on top of Moodle’s VPL plugin. Both resources were made available for 43 classes and 1,407 students. The proposed resources were used by 28 of the classes. A questionnaire sent to the students returned 443 responses, which showed their preferences: Colab (67%), slides (18%) and Moodle activities (15%). 293 assert preferring synchronous lectures. Moreover, around 82% of the respondents approved our method.