Abstract

Over the past decade, the landscape of teaching statistics and data science has been valued in the job market. In response to these trends, academic administrators and educators are increasingly focusing on optimizing learning outcomes and adapting their teaching strategies to serve student needs better.
Our research addresses these shifts by utilizing Natural Language Processing (NLP) to gain deeper insights into student feedback, going beyond traditional survey methods like Likert scales. NLP facilitates a nuanced understanding of student attitudes, interactions, and overall engagement, offering a more personalized feedback mechanism that captures the 'personal' perspectives of students.
The data for this study is gathered from the Israeli academic college, specifically targeting students in non-STEM specializations. These students engage in a statistics course during their 1st or 2nd academic year and a data mining course in their 2nd year. All courses employ a project-based learning methodology, with ongoing projects throughout the course. This practical approach is integral to our research focus.
The goals of our research are multifaceted, aiming to enhance the administration of statistics and data science courses by understanding student preferences and refining teaching methods. We seek to assess the effectiveness of integrating practical elements into theoretical curricula and explore how personalized learning approaches impact student academic achievements and career opportunities. Furthermore, we aim to adapt existing NLP algorithms for educational purposes, specifically for enhancing the teaching of quantitative thinking and data science. This involves developing specialized NLP models that cater to the unique linguistic and contextual nuances of education in these fields.
Methodologically, our study employs a combination of structured and unstructured data collection techniques, including closed and open questionnaires, as well as semi-structured interviews. This approach ensures a comprehensive analysis of both quantitative and qualitative aspects of student feedback. The anticipated outcome is a set of refined educational strategies that are directly informed by student needs and preferences, ultimately leading to improved learning experiences and better preparation for the job market.
By leveraging the capabilities of NLP and embracing the shift towards more practical and personalized education, our research aims to contribute significantly to the evolving field of statistics and data science education, ensuring that graduates are not only well-versed in theory but also adept in practical applications relevant to their future careers.