Leng, D., & von Davier, M. (2025). Gender differences and types of test-taking behavior in PIRLS 2021. Applied Measurement in Education, 1-19. https://doi.org/10.1080/08957347.2025.2486032
Leng, D., Bezirhan, U., Khorramdel, L., Fishbein, B., & von Davier, M. (2024). Examining gender differences in TIMSS 2019 using a multiple-group hierarchical speed-accuracy-revisits model. Educational Measurement: Issues and Practice. 43(3), 64-75. https://doi.org/10.1111/emip.12606
I extended the recently proposed hierarchical speed-accuracy-revisits model to multiple country-by-gender groups to examine the extent to which mathematics ability, response speed, revisit propensity, and the relationship among them differ between boys and girls.
The same first-level models were used for all students and items but different distributions of person parameters for different groups were assumed at the population level. Model estimation used the Bayesian approach.
The model can be used to examine group differences in test-taking behavior not limited to gender.
Leng, D.. How does private tutoring contribute to mathematics achievement in South Korea? An application of hierarchical linear models on TIMSS 2019 data. Fudan Education Forum, 22(3), 68-81. [in Chinese]
I investigated the effects of private tutoring on mathematics achievement based on TIMSS 2019 eighth-grade data using a two-level linear model.
The findings indicate that at the individual level, the longer South Korean eighth graders engaged in tutoring, the better their mathematics scores. On top of that, students' performance was predicted to further improve if their school's average tutoring duration is long; however, this school-level effect could be partly accounted for by the socio-economic background of its students.
The relationship between tutoring and academic achievement did not vary by school or academic level; however, the lower a student's family's economic and cultural capital, the stronger the relationship between extended tutoring and improved scores.
The study also uncovers that the quality of in-school mathematics instruction could moderate the effect of tutoring on achievement: the clearer the student perceived their math teacher's instruction to be, the lesser the association between more than eight months of tutoring and improved scores.
Lu, Y., Leng, D., & Shen, Z. (2024). From winning to serving: Value orientations in the formation of top-notch students’ aspirations. Educational Research. [in Chinese]
How do gifted students prioritize values among themselves, their families, and the society? I used three-step latent class analysis, the Lanza method, and the BCH method to a) identify types of value orientations, and b) study the relationship between the value orientation types and academic achievement, time spent on research, and family socioeconomic status based on a survey of 1,543 students from the top-notch STEM talent cultivation programs of 7 prestigious universities in China.
Four types of students were identified: the "leader", "expert", "utilitarian", and "alienation" types.
Lu, Y., & Leng, D.* (2020). How did advanced learning experiences in secondary school contribute to performance of students in top-notch talent cultivation programs in college?. Peking University Education Review, 18(4), 129-150. [in Chinese]
* Correspondence author
I conducted ordinal Logistic regression, Logistic regression and linear regression based on a survey of 1,359 students of top-notch STEM talent cultivation programs from 6 elite universities in China and found that students with advanced learning experiences (winning a mathematics or science Olympiad award, finishing a scientific research project, or having access to university education resources) in secondary school significantly performed better in college.
These advanced learning experiences in secondary school were significantly related to home and geographic contexts.
Lu, Y., Liu, M., & Leng, D. (2017). The effects of the full implementation of ‘lectures in large classes and discussions in small classes’ on the general education core curriculum. Journal of Higher Education, 38(8), 69-78. [in Chinese]
I supported data analysis from a sample of 1020 undergraduate students from Fudan University and offered recommendations for teachers to provide explicit evaluation criteria in small-class discussion.
The study shows that the “lectures in large classes and discussions in small classes” policy benefited students by deepening their understanding in learning material and enabling them to accept the existence of different views.
Leng, D., Fu, Y., & Han, K. T. (2025, April 23-26). Understanding the behavior and implications of rapid guessing. Annual Meeting of the National Council on Measurement in Education (NCME), Denver, CO, United States.
Leng, D., & von Davier, M. (2024, April 11-14). Using process data to explain gender differences in PIRLS 2021. Annual Meeting of the National Council on Measurement in Education (NCME), Philadelphia, PA, United States.
Leng, D., Bezirhan U., & von Davier, M. (2023, July 24-28). Leveraging process data and variable selection for TIMSS achievement estimation [Paper presentation]. International Meeting of the Psychometric Society (IMPS), College Park, MD, United States.
Leng, D., Khorramdel, L., Bezirhan U., & von Davier, M. (2023, April 12-15). Using process data to improve the estimation of achievement in TIMSS 2019. In L. Khorramdel, & M. von Davier (Chair), Advanced Technology Use in TIMSS and PIRLS [Symposium]. Annual Meeting of the National Council on Measurement in Education (NCME), Chicago, IL, United States.
Leng, D., Zhu, M., & Lu, Y. (2022, April 21-26), Latent class analyses of students’ attitudes toward mathematics in the U.S., Singapore, and Korea [Poster presentation]. Annual Meeting of the American Educational Research Association (AERA), San Diego, CA, United States.
Leng, D. (2022, January 15-16), How does private tutoring contribute to mathematics achievement? Two-level linear modeling on TIMSS 2019 South Korean data [Paper presentation]. The First International Doctoral Student Forum on Educational Research at Fudan University, Shanghai, China. [in Chinese]
Leng, D., Bezirhan U., Khorramdel, L., Fishbein, B., & von Davier, M. (2021, November 16-18), Examining gender differences in TIMSS 2019: The multiple group hierarchical speed-accuracy-revisits model [Paper presentation]. 9th International Association for the Evaluation of Educational Achievement (IEA) International Research Conference (IRC), Dubai, United Arab Emirates.
Lu, Y., Liu, M., & Leng, D. (2017, November 6) A case study of the reform of undergraduate teaching at Fudan University [Paper presentation]. Graduate Student Academic Forum at East China Normal University, Shanghai, China. [in Chinese]
Leng, D., & Luo, L. (2017, October 13), A comparison of the questionnaire used in “Assessment of General Education Core Curriculum at Fudan University’ and 11 other assessment tools [Presentation]. Symposium on the Reform of General Education at Fudan University, Shanghai, China. [in Chinese]
“Assessment of general education core curriculum at Fudan University (2016),” General education core curriculum committee and Research Institute for Higher Education of Fudan University (submitted to Fudan University, with Yi Lu, Wei Xu, Shan Lin, Tingting Zhang, and Wangshurui Chen)
“Assessment of general education core curriculum at Peking University (2016),” Research group of general education core curriculum at Peking University (submitted to Peking University, with Yi Lu, Tianhui Huang, Yu Cao, and Li Dong)
In an effort to improve the overall quality of general education in China’s top universities and to assess the learning gains of students, the research group conducted a large-scale periodic student survey at Fudan University and Peking University. The effective samples of the two reports were 3633 and 1684, respectively. Based on pre-survey results, we developed 3 indicators and 20 sub-indicators in evaluating the quality of the courses.