Table 6 Estimates of the presence of at least one SEN student on their peers’ achievement, by different characteristics.

From: Students with special educational needs in regular classrooms and their peer effects on learning achievement

 

2P

2S

Reading

Mathematics

Reading

Mathematics

Coef.

Std. Err.

Obs.

Coef.

Std. Err.

Obs.

Coef.

Std. Err.

Obs.

Coef.

Std. Err.

Obs.

Student’s gender

Male

0.005

(0.027)

29,003

0.004

(0.033)

28,989

0.013

(0.024)

24,390

0.026

(0.027)

24,384

Female

0.029

(0.027)

26,634

0.029

(0.033)

26,625

−0.010

(0.024)

23,101

0.031

(0.029)

23,100

 Clogg’s z-test (p-value)

−0.629 (0.530)

−0.536 (0.592)

0.678 (0.498)

−0.126 (0.900)

Mother’s age

Mean age or younger

−0.009

(0.027)

28,486

−0.017

(0.033)

28,471

0.011

(0.023)

24,526

0.048*

(0.027)

24,519

Over than mean age

0.039

(0.026)

27,151

0.043

(0.034)

27,143

−0.017

(0.024)

22,965

0.003

(0.028)

22,965

 Clogg’s z-test (p-value)

−1.281 (0.200)

−1.266 (0.205)

0.842 (0.400)

1.157 (0.247)

Mother’s education

No tertiary education

0.017

(0.023)

50,775

0.022

(0.030)

50,755

0.005

(0.018)

45,052

0.031

(0.022)

45,046

Tertiary education

0.014

(0.077)

4862

−0.061

(0.075)

4859

−0.095

(0.102)

2439

−0.043

(0.106)

2438

Clogg’s z-test (p-value)

0.037 (0.970)

1.028 (0.304)

0.965 (0.334)

0.684 (0.494)

Cohort size

Between 10 - 20 students

0.022

(0.038)

28,174

0.029

(0.048)

28,161

0.043

(0.029)

23,862

0.052

(0.034)

23,858

Between 21 - 30 students

−0.023

(0.035)

27,463

−0.035

(0.045)

27,453

−0.034

(0.029)

23,629

0.010

(0.037)

23,626

 Clogg’s z-test (p-value)

0.871 (0.384)

0.973 (0.331)

1.877 (0.060)*

0.836 (0.403)

  1. Each coefficient represents a separate regression. The median age in 2P is 41.5 years and in 2S is 44.8 years. All the regressions include year and school fixed effect as well as student, cohort, school, and family controls. Student controls include sex, language, and age. Cohort controls include mean age, proportion of male students, proportion of students who speak an indigenous language, and number of students. School controls include the number of students. Family controls include mothers’ age and education and number of siblings. Standard errors clustered at the school level were used to estimate each coefficient.
  2. ***p < 0.01; **p < 0.05; *p < 0.10.
Back to article page