Gender differences between teachers’ assessments and test based assessments. Evidence from Spain

ABSTRACT

We study the existence of a gender gap between two methods of grading student achievement in Spain, namely teachers’ assessment which measures cognitive outcomes and standardised test scores, which are consistent across Spanish schools. We use rich census information of Andalusian students attending secondary schools. Our results indicate that overall teachers’ assessment favours girls compared to the result of standardised test scores in reading and maths. On average, girls obtain 0.35 points more than boys in teacher’s assessment on a 1 to 10 scale (0.15 SD), conditional on both having the same competencies. Additionally, we have found that the gender gap is stable across different levels of students’ own ability and peers’ ability. These results are robust to a set of census social characteristics. The presence of grading disparities suggests that test scores could complement teachers’ assessment; specifically, they could be used to inform about the gender gap in assessments.

KEYWORDS:

• Teachers’ assessment
• test scores
• skills
• knowledge
• census

Acknowledgments

We are grateful to the Consejería de Educación of the Junta de Andalucía for data support. This work has been partly supported by FEDER funding (under Research Project PY20-00228-R); Ministerio de Ciencia e Innovación (under Research Project PID2020-119471RB-I00); the Andalusian Regional Government (SEJ-645).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/0969594X.2023.2251715

Notes

1. For instance Mostafa (2019) and Breda and Napp (2019) suggest that a potential factor is that women perform relatively better in reading compared to maths and science.

2. Note that Copur-Gencturk et al. (2020) find that teachers’ assessment stressed gender stereotypes by subject (Copur-Gencturk et al. 2020).

3. Hay and Macdonald (2008) in Australia; Lavy (2008) in Israel; Burgess and Greaves (2013) and Campbell (2015) in England; Rangvid (2015) in Denmark; Diamond and Persson (2016) in Sweden; Calsamiglia and Loviglio (2019) in Spain; Terrier (2020) in France.

4. Programme for International Students Assessment.

5. The average academic performance in PISA 2018 was: a) In reading 477 score points in Spain and 466 in Andalusia; b) In maths 481 in Spain and 467 in Andalusia (authors’ own calculation).

6. Authors’ own calculation based on PISA 2018 data for Andalusia.

7. The scale of school climate consisted of eight domains: student – teacher relationships school connectedness academic support order and discipline school physical environment school social environment perceived exclusion/privilege and academic satisfaction.

8. Orden de 27 de octubre de 2009 de la Consejería de Educación (Junta de Andalucía) por la que se regulan las pruebas de Evaluación de Diagnóstico y el procedimiento de aplicación en los centros docentes de Andalucía.

9. Ley Orgánica 2/2006 de 3 de mayo de Educación.

10. In reading and maths Hausman’s test rejects the null hypothesis at 1% (p-value = 0). The ${\chi }^2$ statistic is equal to 446.29 in reading and 282.91 in maths.

11. If a student does not attend school on the test date they cannot repeat the test.

12. To the extent that attendance is compulsory those who do not pass the test are a priori a random selection of the sample because of any health related issue.

13. Average standard scores of the sample included (not included) in the estimations are: 0.023 (−0.021) in test scores in reading; 0.018 (−0.154) in test scores in maths; 0.073 (−0.457) in teachers’ assessment in reading; 0.069 (−0.436) in teachers’ assessment in maths.

14. 1 SD increase in maths would increase approximately 2.28 points on a scale of 1 to 10.

15. 1 SD increase in reading would increase approximately 2.23 points on a scale of 1 to 10.

16. Hypothesis testing in Table 1 Specification 4 a) Reading. Test for differences between coefficients rejects the null hypothesis: $H_0$: Coefficient of test scores in reading = Coefficient of class-mean test score in reading. The p-value of the test is equal to 0.0000.

17. Hypothesis testing in Table 1 Specification 4 b) Maths. Test for differences between coefficients accepts the null hypothesis: $H_0$: Coefficient of test scores in maths = Coefficient of class-mean test score in maths. The p-value of the test is equal to 0.1387.

18. An alternative set of estimations has explored if the difference in scores could be considered the difference in relative position. In this case, the dependent variable is created on the differences in quartiles of TA and TS distribution. We do not identify any significant change. Results are available upon request to the authors.

19. In public schools, teachers’ selection is based on state exams, while for employment in semiprivate schools’ the selection process is not based on the state exams, although that does not mean that candidates are not selected according to merit and capability.

Additional information

Funding

The work was supported by the Ministerio de Ciencia e Innovación [Research Project PID2020-119471RB-I00]; Andalusian Regional Government [SEJ-645]; FEDER funding [Research Project PY20-00228-R].

Notes on contributors

Oscar D. Marcenaro-Gutierrez

Oscar D. Marcenaro-Gutierrez is a Full Professor at the Department of Applied Economics (Statistics and Econometrics, University of Malaga). He moved to the University of Malaga after working three years at the London School of Economics. He has been working on projects focused on economics of education and labour market. His research interests specially include: education production functions, efficiency, time use, teachers' satisfaction and working-life balance. He has published more than 100 papers, most of them in journals with high impact index. At present he works for different national and international evaluation agencies.

Claudia Prieto-Latorre

Claudia Prieto-Latorre is a post-doctoral researcher in the Faculty of Economics and Businesses at University of Malaga. Her research interests focused on Economics of Education, with an emphasis on peer effects and school compositional effects as well as measures of alternative educational outcomes.

Maria Isabel Sánchez Rodriguez

María Isabel Sánchez Rodríguez is assistant professor on Statistics and Operations Research at the University of Cordoba. Her research focuses on the application of statistical tools to the evaluation of social sciences problems. In particular, she has worked on projects related to organised crime and the rights of ethnic groups, with special focus on gender issues.