Abstract:

This paper explores the growing importance of big data analytics in organizations and the resulting demand for skilled data scientists. The authors note that many educational institutions have responded to this demand by offering formalized data science degrees and curricula, but there is a mismatch between the skills provided in these programs and industry expectations.

The COVID-19 pandemic has disrupted traditional education, forcing schools to transition to online learning. Schools closed during the first wave in Portugal, and courses were moved online. As time passed and multiple waves occurred, hybrid emergency approaches were implemented, allowing students to attend classes online or in person. However, face-to-face attendance decreased throughout the semester as online classes were more convenient and comfortable.

To mitigate this issue and promote the development of highly demanded skills, we propose implementing an active learning approach using Team-Based Learning (TBL) in a Machine Learning and Data Science course. We outline the benefits of TBL, also required by the industry, including promoting active learning, engaging more students in interactive classes, developing personal relationships, promoting group learning, and developing critical thinking skills. We note that TBL provides dynamic learning through group interactions and can be particularly effective in a Machine Learning and Data Science course, where theoretical and practical components exist.

We proposed some modifications to the standard TBL approach. A survey was used to match students' technical skills and personality traits to create diverse and more effective groups. TBL was applied in both theoretical and practical classes, allowing students to apply concepts learned through Python exercises. Personalized support was provided to students through a global dashboard that tracked their quiz results, allowing professors to provide targeted help.

The results indicate high overall student satisfaction after introducing Team-Based Learning (TBL). Students found TBL to be beneficial for their personal and professional development. We have analyzed the results from student assessments, individual and team tests, grades for the final project, and final exam grades. We also surveyed students and performed personal interviews to capture students' feedaback. We showed that our approach also promotes general performance improvement among students.

Keywords:

Data science, hybrid learning, Team-Based Learning (TBL), performance improvement.