Lingnan’s AI transformation in research
Endeavouring to integrate innovation with its long-standing tradition of liberal arts education, Lingnan University was one of the first institutions in Hong Kong to adopt an open approach towards artificial intelligence (AI) in learning and teaching. Not only are staff and students allocated a monthly quota of ChatGPT and encouraged to use AI to reinforce research, but senior management and faculty also take a positive attitude towards incorporating AI in instructions and analysis, and organise topical AI workshops and seminars. In alignment with the University’s guiding motto of “Education for Service”, this initiative of making ChatGPT available throughout the university community helps reduce barriers to entry and inequalities due to differences in financial resources.
Nie Cong, a tutor and PhD student of the Department of Economics, having thought about the possibilities of AI in terms of data analysis and its open network of information, saw an opportunity to address the need for AI transformation research through the statistical software Stata. Known for its openness and the many commands developed externally by its user community, Stata is taught globally in many undergraduate and postgraduate programmes across disciplines. For instance, most of its data processing formulas and functions are supported by a community of scholars for users to install.
Nie Cong, a tutor and PhD student of the Department of Economics at Lingnan University, and some collaborators integrate emerging AI technology and statistical analysis by embedding a new ChatGPT command in Stata.
Nie, who frequently uses the software herself, had the idea of developing a widely used command to enhance the user experience of the software. With collaborators, she successfully assimilated AI capabilities to statistical software by embedding ChatGPT in the programme. This new ChatGPT command, developed to bridge the gap between statistical analysis and burgeoning AI technology, not only supports in-context learning, but also offers an interface that allows Stata users to interact directly with GPT (Generative Pre-trained Transformer) responses within the software itself. By making advanced analytics more accessible to researchers in social sciences, this function will lighten their workload and shorten academic learning time.
One of Stata’s users, Assistant Professor of the Department of Sociology and Social Policy Prof Francisco Olivos, compares GPT’s work to programming assistants, to figure out the problem of codes, make them more efficient, and even locate a specific code for what is conceptually known. As an educator, he finds AI tools beneficial to teaching and learning, and especially in relieving students’ statistical anxiety by offering comprehensive support in their studies. “Instead of waiting for an email response from a human tutor or their instructor, AI provides 24/7 assistance with even more knowledge. The package developed by our student will definitely have a positive impact on statistical courses at Lingnan and beyond,” Prof Olivos says.
AI also serves as an invaluable tool to expedite the data handling processes of Nie’s current research, which focuses primarily on international trade and patent applications. Countries are often named differently across databases, for example “Mainland China” and “China (the Mainland)”, and the ChatGPT command generates uniform country names in ISO3 form, reducing the time spent on data processing and minimising errors that might occur in manual operations.
It is crucial to evaluate the integration of research methodologies with AI applications.
Nie also suggests offering guidance and financial support to students who can expand this application within their specific academic fields. "Such support would motivate students to delve deeper into AI, thereby enhancing their research outputs and potentially leading to innovative developments in various disciplines," she says, adding that her team will continue to update the command and develop more user functions.