Do teacher and classroom characteristics affect the way in which girls and boys are graded?

Abstract

Teachers’ evaluations of students do not consider only academic competence, but are imbued with social considerations related to individual teacher and student characteristics, their interactions, and the surrounding context. The aim of this paper is understanding the extent to which teachers grade girls more generously than boys, and which characteristics of teachers and classrooms are likely to reduce this gender grading gap. We use Italian data from INVALSI-SNV, providing information on 10th-grade students linked with their teachers. The analysis relies on grade equation models in multilevel regression analysis, with students as first level, teachers/classrooms as second level, and schools as third level. Results show that, when comparing students who have identical subject-specific competence, teachers are more likely to give higher grades to girls. Furthermore, they demonstrate for the first time that this grading premium favouring girls is systemic, as teacher and classroom characteristics play a negligible role in reducing it.

Keywords:

• Teachers’ grades
• gender
• grading mismatch
• academic performance
• education
• social inequalities

Disclosure statement

The authors report there are no competing interests to declare.

Notes

1 Calculation based on www.dati.Istat (Italian national statistical office), 2019 data.

2 Miur: Ministero dell’Istruzione, dell’ Università e della Ricerca, divided in 2020 into two different ministries: Ministero dell’ Istruzione (Ministry of Education) and Ministero dell’ Università e della Ricerca (Ministry of University and Research).

3 https://www.miur.gov.it/valutazione

4 https://www.miur.gov.it/web/guest/patto-educativo-corresponsabilita

5 CAWI: Computer Assisted Web Interviewing

6 For more information about INVALSI test score: www.invalsiopen.it

7 Index provided from INVALSI that measures students’ economic, social and cultural status. It is a synthesis of three indicators: (1) Parental occupational status; (2) Parental level of education; (3) Possession of specific material assets.

8 We tested the model fit introducing also age squared. Both Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) suggest that age squared does not improve the model fit.

9 The expression “within-school seniority” refers to the length of teacher’s experience within the school.

10 Medium-high and higher ESCS students correspond to the 3rd and 4th quartile of ESCS distribution

11 Another possibility would be to assess the teacher grading premium with a difference-in-difference design (Di Liberto, Casula and Pau 2021). Teacher grading premium can be understood as the difference between girls and boys, gathered via the difference between teacher grade (‘non-blind’ measure) and the standardized test score (‘blind’ and supposedly unbiased measure) (see Lavy 2008). Consequently, teacher grading mismatch would be defined as the average gap between for female students, minus the same gap for male students, as prior researchers have done to estimate gender discrimination (Falch and Naper 2013; Goldin and Rouse 2000). However, in the current context, this would also be feasible by standardizing the different scores in order to compare them.

12 It is important to note that these models also include controls for classroom composition. Therefore, our estimates regard GGG for female and male students in different types of schools that have similar classroom composition. Estimates without controls for classroom composition suggest similar patterns.

13 INVALSI-SNV data does not collect and provide such information.

Additional information

Funding

This work was supported by Compagnia di San Paolo.