CACLER 20th Anniversary Knowledge Exchange Series on machine learning applications in PISA data analysis

The online seminar entitled ‘Influencing Factors on Resilient Students: A Study Based on PISA 2018 Data through Machine Learning Confirmation’ delivered by Prof. Zhao Ningning, Professor, Beijing Normal University, was well completed on 21 March 2024. Many thanks to Dr. WI Lam who is the Co-investigator of PISA 2025 served as the discussant provided insightful responses after the seminar. The seminar was attended by 39 persons online. Most of them are University students while some of them are teachers and research staffs. 12 of them have responded to the online questionnaires and most of them found that the seminar is useful (4.92 out of 6 at max. in average) and inspiring (4.83 out of 6 at max. in average). Many participants found the research method the most useful which can be applied in their research study in future. Yet, they thought more discussions and Q&A on applications/demonstrations to machine learning are needed after the presentation.

Chair: Prof. LIN Chin-Hsi, Director of CACLER; Associate Professor, Faculty of Education, HKU

Discussant: Prof. LAM Wai Ip, Co-investigator of PISA 2025; Associate Professor, Faculty of Education, HKU

This study explored the factors influencing the academic resilience of disadvantaged students using PISA 2018 data. The Boruta algorithm was employed to identify key factors, while XGBoost was used to predict students’ resilience in reading, math, and science. The Shapley Additive Explanations (SHAP) framework was utilized to estimate the impact of significant factors. The results showed five common factors among the top ten in the three subjects, including sense of meaning in life, classroom disciplinary climate, perception of cooperation at school, expected career status, and fear of failure. Variations in other factors were observed across the different resilience domains. This study offers fresh insights into academic resilience and highlights the critical need to nuture academic resilience among disadvantaged students.

Interested parties please register online (Click HERE～HERE～) on a first-come, first-served basis no later than 20 March 2024.

1) Confirmation email (including meeting link) will be sent to the registered email addresses once approved.

2) Participants are required to click the meeting link in the email message to join.

(Online Workshop) Benefits and Limitations: Machine Learning Applications in PISA Data Analysis

The online workshop entitled ‘Benefits and Limitations: Machine Learning Applications in PISA Data Analysis’ delivered by Prof. Zhao Ningning, Professor, Beijing Normal University, was well completed on 26 March 2024. Mr. Biao Zeng, PhD candidate supervised by Prof. Zhao, had provided a thorough demonstration on setup procedures in using the R and Jupyter to run data. The workshop was attended by 24 persons online. Most of them are University students while some of them are teachers and research staffs. Six of them had responded to the online questionnaires and most of them found the workshop both informative and inspiring (5.17 out of 6 at max. in average). Overall speaking, they are highly satisfied with the workshop (5.33 out of 6 at max. in average). Some of them found application of R and Jupyter and use of code the most useful. One responded that he/she can now conduct analysis of large quantity of data after this workshop while another would like to learn more details on each line of code.

Chair: Prof. LIN Chin-Hsi, Director of CACLER; Associate Professor, Faculty of Education, HKU

Machine Learning (ML) techniques have revolutionized various fields, demonstrating their effectiveness in applications such as data mining for the Programme for International Student Assessment (PISA). This workshop aims to delve into the benefits and limitations of ML methods. It will showcase the impact achieved through the fusion of ML and education while addressing the remaining challenges. Attendees will engage in running ML experiments using Python and participate in discussions to gain a comprehensive understanding of the subject. The workshop seeks to explore challenges, propose potential solutions, and chart future directions for Machine Learning in education.

Interested parties please register online (Click HERE～HERE～) on a first-come, first-served basis no later than 25 March 2024.

1) Confirmation email (including meeting link) will be sent to the registered email addresses once approved.

2) Participants are required to click the meeting link in the email message to join.

Ningning Zhao is a Professor in the Department of Chinese Language Educational Studies at Beijing Normal University. With expertise in educational assessment and research, she holds a Ph.D. in Curriculum and Instruction. Her research interests include curriculum development, instructional strategies, and quantitative analysis in social science. Prof. Zhao has actively participated in various research projects, including textbook evaluation and education quality assessment. Her commitment to improving educational practices is evident through her diverse experiences and involvement in social work. For details about the speaker and her publications, please visit:

Language and Literacy Education (LALE) Unit, Faculty of Education, HKU
The Consortium for Research on Language Policy and Practice (CRLPP)

CACLER 20th Anniversary Knowledge Exchange Series on machine learning applications in PISA data analysis

(Online Seminar) Influencing Factors on Resilient Students: A Study Based on PISA 2018 Data through Machine Learning

(Online Workshop) Benefits and Limitations: Machine Learning Applications in PISA Data Analysis