lead in math achievement compared to their female peers throughout primary education. Using the IV estimates, we conclude that female students reduce their initial gap with their male peers in math skills, although they end up with a lower skill level at the end of primary education. Given our notion of the bias in OLS estimates, we will rely on the IV estimates. Then, we calculate the proportion of the sixth-grade achievement gap that is formed before and within primary education. For reading, we conclude that around 18% of the sixth-grade achievement gap between male and female students is formed during primary school. The remaining 82% of this gap was already present before second grade. For spelling, we would obtain a negative percentage of the sixth-grade achievement gap that is formed during schooling, which is the reason why we do not report this percentage. This, however, means that this achievement gap is completely produced before second grade. Also for math, the complete sixth-grade achievement gap is formed before second grade, considering the IV estimates. Note, however, that we would conclude that around 12 percent of this achievement gaps if formed between second and sixth grade if we would consider the OLS estimates. Figure B.4 of the appendix examines in which grades disparities between male and female students are mainly formed throughout primary education. For reading, the total learning advancement difference between male and female students throughout primary education remains rather stable, albeit that there is an outlier in third grade. This means that, despite this outlier, the proportion of the sixth-grade achievement gap that arises within the primary school period remains 71
RkJQdWJsaXNoZXIy MjY0ODMw