Table B.13: Robustness check of the sixth-grade achievement gap with imputed value for parental education Parental education (Middle=ref.) Sex (Male=ref.) Migration background (None=ref.) Low High Female Western Non-Western Panel I: Reading Grade 6 -.146*** .365*** .239*** .077*** -.029 Grade 6, control grade 2 OLS -.089*** .240*** .136*** .092*** .056*** (60.96) (65.75) (54.90) (119.48) / IV -.061*** .169*** .040*** .128*** .167*** (41.78) (46.30) (16.73) (166.23) / Panel II: Spelling Grade 6 -.059*** .194*** .045*** .142*** .156*** Grade 6, control grade 2 OLS -.028 .130*** .000 .179*** .209*** (47.46) (67.01) / (126.06) (133.97) IV -.018 .067*** -.089*** .111*** .146*** (30.50) (34.53) / (78.17) (93.59) Panel III: Math Grade 6 -.114*** .351*** -.276*** .126*** .013 Grade 6, control grade 2 OLS -.075*** .220*** -.128*** .139*** .172*** (65.79) (62.67) (46.38) (110.32) / IV -.056*** .145*** .045*** .100*** .178*** (49.12) (41.31) / (79.37) / Note. This table presents the estimates for the regressions of sixth-grade skill scores on student characteristics (parental education, age, sex, migration background, household type, and family size), grade retention, a covid dummy, and year FE. We have imputed missing values for parental education, where feasible. In the first row of each panel, the estimate for the corresponding student characteristic is reported. In the second and third rows, the second-grade skill score is added to the regression, and the updated estimates for the student characteristic are shown for the OLS and IV regressions, respectively. The proportion of the sixth-grade achievement gap that forms between second and sixth grade is displayed in parentheses. A forward slash indicates that the learning advancement of a specific group is in the opposite direction of the sixth-grade achievement gap. This would result in a negative proportion of the sixth-grade achievement gap formed between second and sixth grade, which we do not report. A forward slash is also displayed if the sixth-grade achievement gap is not significant. *** p< .01, ** p< .05, * p< .1. 217
RkJQdWJsaXNoZXIy MjY0ODMw