In this project, we compare the students’ results at the end of Secondary Education in two different types of assessment: teachers’ grading and scores on national exams.
There has been a discussion regarding the weight that should be attributed to each of these types of assessment, for example in what concerns access to Higher Education, and there are even those who propose the elimination of national exams altogether. It is relevant to understand the implications of these different options.
In a previous work, we had already seen that when we compare exams and teacher-assigned grades, on average and in relative terms, boys are favoured in exams and girls in the grades assigned by their teachers.
This implies that, if we increase the weight of exams in the selection process to access Higher Education, probably more boys will be enrolled, and if we eliminate national exams, then probably the existing gender imbalance will increase.
This is something that we had already observed in Portugal, as well as in other countries.
In the current work, we add an analysis on whether this difference in the behaviour of boys and girls in different assessment methods could be explained by teacher’s gender.
We know that most teachers are women, and we thought that this could be an explanation for these results.
To study this topic, we use administrative data on the Portuguese education system for more than 20,000 teachers and 400,000 students that took the national exams in their 11th and/or 12th grades in 12 different subjects, which covers all areas, between 2008/09 and 2016/17.
Our identification strategy is based on the fact that we compare grades for the same student and in the same subject, which allows us to control for the effect of individual student characteristics not observable by the researcher. Additionally, the fact that we have the same student taking several subjects, in some cases with male teachers and in others with female teachers, also allows us to control for a possible bias in the allocation of students to teachers. This allows us to describe the results obtained as causal effects of the type of assessment and the teacher’s gender on the student’s grade.
We conclude that, both with male and female teachers, on average and in relative terms, it is better for boys to be assessed through exams and for girls by their teachers (male or female).
Being more specific: we observe that students, boys and girls, tend to have lower grades on exams than those given by their teachers, but also that:
- This difference is lower for boys (grades fall less in exams), both with male and female teachers.
- This difference is higher for male teachers (grades fall more in exams).
Conclusion: when we think about changing the weights of the different components in the access to Higher Education, there might be unforeseen and unwanted consequences.
It might be interesting to think about different solutions as for example the one implemented in Sweden, in which students can choose different forms of access, based on different criteria, each with their specific pre-set number of available slots. Thus, students with different characteristics can choose the process which most values their competences.
Read the paper here: https://dx.doi.org/10.2139/ssrn.4048290