Table 1.
Cohorts of students analyzed.
Table 2.
Number of twins per cohort.
Table 3.
Descriptive statistics of the student population, total population and twins population, 2015–2020 cohorts.
Fig 1.
Each dot represents a student according to their 4th grade Language SIMCE (x-axis) and Mathematics (y-axis) score. Both scores are measured in units of standard deviation. The students were classified into groups according to both scores, using the k-means methodology with Euclidean distance; the dot color shows the group each student is in.
Table 4.
Models to predict Mathematics performance in competitive (PSU) and noncompetitive (10th grade SIMCE) tests.
Fig 2.
Predictions of Mathematics scores in competitive (PSU) and noncompetitive (10th grade SIMCE) tests.
The figure shows predictions for students with average characteristics but different gender (panel A), different previous Mathematics SIMCE test scores and gender (panel B), and different achievement groups and gender (panel C). All the predictions show 95% confidence intervals. The estimations of panels A, B, and C were made with models (1), (2), and (3), respectively. Table 4 shows the coefficients of those models.
Table 5.
Gender gaps comparison for different levels of previous achievement.
Estimations based on Model (3) for Mathematics.
Fig 3.
Predictions of Language scores in competitive (PSU) and noncompetitive (10th grade SIMCE) tests.
The figure shows predictions for the students with average characteristics and different gender (panel A), different previous Language SIMCE test scores and gender (panel B), and different achievement group and gender (panel C). All the predictions show 95% confidence intervals. The estimations for panels A, B, and C were made with models (1), (2), and (3), respectively. Table 6 shows the coefficients of these models.
Table 6.
Models to predict the performance in language competitive tests (PSU) and noncompetitive tests (10th grade SIMCE).
Table 7.
Gender gaps comparison for different levels of previous achievement.
Estimations based on Model (3) for Language.
Fig 4.
School grades growth between 10th and 12th grade by gender.
Panel A shows the growth in Mathematics grades between 10th and 12th grade in men and women. Panel B shows the growth in Language grades between 10th and 12th grade in men and women. The grades are measured in standard deviation units, and the growth is calculated as 12th grade scores minus 10th grade scores.