Across many education systems, the recruitment of mathematics teachers relies heavily on two easily observed credentials: teachers’ highest education level and subject major (Ingersoll, 2007). These credentials are routinely used as screening criteria in hiring and placement decisions because they are administratively simple and assumed to reflect professional capacity. However, there is no shared or consistent standard for how these two indicators should be interpreted or weighted in practice. Even within the same education system, requirements and preferences can differ across regions and local authorities (Ma et al., 2021). For example, regions may differ in how strictly “major matching” is enforced, whether a Master’s degree is treated as an advantage or a requirement, and whether certain combinations of education level and major are prioritised for particular grades or schools. This creates uncertainty about what these credentials actually capture and whether they are consistently linked to students’ mathematics outcomes.

This paper uses international large-scale assessment data to examine how teachers’ formal qualifications relate to individual students’ mathematics achievement. It brings together two TIMSS cycles that capture different stages of mathematics learning: TIMSS 2023 Grade 8 (lower-secondary) and TIMSS Advanced 2015 (students in the final year of secondary schooling who have taken or are taking advanced mathematics, typically around age 18). Using a common modelling strategy across the two datasets, the study examines whether the same teacher qualifications matter in similar ways across stages and whether the combination of teachers’ education level and subject preparation is especially relevant for advanced mathematics. The main aim of this paper is to provide comparative evidence that can help inform mathematics teacher training and entry requirements, and support stage-specific staffing decisions. It does not make causal claims, but it uses international data to examine whether teacher education level and subject major are consistently associated with students’ mathematics achievement.

Research questions

1. Are students’ mathematics achievement scores associated with their teacher’s highest education level and subject major, after accounting for student background and school context?

2. Is there an interaction between teachers’ highest education level and subject major in relation to students’ mathematics achievement?

3. Are these patterns similar between lower-secondary mathematics (TIMSS 2023 Grade 8) and advanced mathematics at the upper-secondary level (TIMSS Advanced 2015)?

This study is a secondary analysis of international large-scale assessment data from TIMSS 2023 (Grade 8) and TIMSS Advanced 2015 (advanced mathematics, final-year secondary). The two datasets are analysed separately, but the modelling strategy is kept as consistent as possible across them by using parallel outcome definitions and a matched set of student-, school-, and teacher-level variables available in both cycles. The unit of analysis is the individual student, linked to their mathematics teacher through the TIMSS student–teacher–school data structure. Student mathematics achievement is the outcome variable. Teachers’ highest education level was coded into four categories: below Bachelor’s, Bachelor’s, Master’s, and Doctorate.TIMSS includes a wide range of specific teacher majors, so teacher subject background was coded using several alternative groupings. The major classifications focus on distinctions between mathematics, mathematics education, and broader education-related fields (including mathematics education within the education category in some specifications). These alternative classifications are implemented through multiple models to examine whether conclusions are sensitive to how subject background is grouped. The analysis has two parts. First, effect sizes are used to describe achievement differences across teacher categories. Cohen’s d is calculated to summarise the magnitude of differences between students taught by teachers with different highest education levels and different subject-major groupings. Second, the main analyses use multiple linear regression with a three-block structure. Block 1 includes student- and school-level background controls and teacher experience. Block 2 adds teacher subject major, operationalised using multiple alternative classifications. Block 3 adds teachers’ highest education level. For each dataset, the study estimates a set of seven models that vary the teacher-major grouping and related specifications, allowing comparison of coefficients and incremental explanatory power (changes in R²) across model variants. The final specifications include interaction models testing whether the association between teachers’ education level and student achievement varies by teachers’ subject major, using education level × major interaction terms.

The results show a broadly consistent pattern for teachers’ highest education level, but the relationship is not simply linear. Students taught by teachers with a Master’s degree tend to achieve higher in mathematics with student and school background while teachers with a Doctorate degree does not appear to bring a further advantage. By contrast, teachers’ subject majors show a clear stage-specific pattern. In Grade 8, education-related majors are generally associated with higher achievement, with mathematics education showing the strongest positive association. In the same stage, teachers with mathematics-only majors perform least well, and the combined mathematics-and-education group (Both) is also associated with lower achievement once background factors are taken into account. In TIMSS Advanced, the pattern shifts: the advantage linked to education-related majors becomes weaker, while mathematics-focused majors show a positive association. A plausible interpretation is that lower-secondary mathematics is not necessarily demanding in terms of advanced content, but it can be demanding to teach well. Helping adolescents grasp basic ideas, correct misunderstandings, and stay engaged often depends on clear explanations and effective teaching strategies. By the advanced stage, students are typically more academically selected and already have stronger foundations, so deeper disciplinary preparation may align more closely with the content demands of high-level courses.This contrast suggests that recruitment and staffing criteria may need to be more explicit about the intended teaching stage, rather than assuming that the same credential profile fits both lower-secondary and upper-secondary level. Finally, interaction models provide little evidence of a meaningful combined effect between highest education level and subject major. Any differences across major groups in the association of education level with achievement are small after controls. Overall, these findings describe robust associations rather than causal effects, but they provide comparative evidence to inform discussion about qualification standards and stage-sensitive workforce planning in mathematics.

Ma Yungjun., Li Li., Ru Zhiwin & Li Danhui. （2021） History and reality: review and prospect of primary and secondary school teacher recruitment. Journal of Guangdong University of Education (06),79-89. Ingersoll, R. (2007). A Comparative Study of Teacher Preparation and Qualifications in Six Nations. 47. https://repository.upenn.edu/handle/20.500.14332/8401