Table 3.
Balance Check of Baseline Characteristics of the Remaining Sample Across Experimental Group
| Variables | (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | (9) | (10) | (11) | (12) | (13) | (14) | (15) | (16) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Age (Year) | Grade 5 (1-Yes) | Male (1=Yes) | Baseline Vison Knowledge (0–8) |
Myopia Knowledge (0–4) | Eyeglasses Usage Knowledge (0–4) | Owns Eyeglasses (1=Yes) |
Baseline Aware of Being Myopic (1=Yes) |
Living at School (1=Yes) |
Baseline Visual Acuity of Worse Eye (LogMAR) | Baseline Standardized Math Scores | Father Has High School Education or Above (1=Yes) |
Mother Has High School Education or Above (1=Yes) |
Both Parents Migrated for Work (1=Yes) |
Household Wealth (Index) | Province (1=Gansu) | |
| Panel A: Balance between Control and Education group accounting for attrition in follow-up | ||||||||||||||||
| Education | 0.018 | −0.031 | −0.053 | −0.146 | −0.088 | −0.059 | −0.048* | −0.019 | −0.036 | 0.050** | 0.057 | −0.025 | 0.025 | 0.044* | −0.069 | 0.026 |
| (0.115) | (0.033) | (0.032) | (0.131) | (0.075) | (0.094) | (0.028) | (0.047) | (0.071) | (0.022) | (0.102) | (0.027) | (0.025) | (0.022) | (0.093) | (0.119) | |
| Attrition between baseline and evaluation | −0.170 | −0.164* | −0.081 | 0.158 | 0.143 | 0.015 | 0.139 | 0.115 | 0.211** | 0.023 | 0.062 | 0.083 | 0.116 | 0.181** | 0.058 | −0.278*** |
| (0.266) | (0.097) | (0.121) | (0.467) | (0.276) | (0.265) | (0.087) | (0.115) | (0.083) | (0.064) | (0.224) | (0.094) | (0.096) | (0.087) | (0.212) | (0.080) | |
| Interaction_attr1 | 0.306 | 0.197 | −0.166 | 0.058 | 0.072 | −0.014 | −0.175 | −0.329** | −0.029 | −0.057 | −0.871** | 0.006 | −0.072 | 0.245 | −0.431 | 0.105 |
| (0.577) | (0.189) | (0.175) | (0.638) | (0.390) | (0.366) | (0.115) | (0.151) | (0.149) | (0.080) | (0.425) | (0.147) | (0.135) | (0.167) | (0.333) | (0.143) | |
| Constant | 10.539*** | 0.614*** | 0.531*** | 2.392*** | 1.257*** | 1.135*** | 0.161*** | 0.464*** | 0.239*** | 0.622*** | 0.204** | 0.167*** | 0.084*** | 0.069*** | 2.192*** | 0.378*** |
| (0.064) | (0.022) | (0.021) | (0.079) | (0.057) | (0.046) | (0.022) | (0.029) | (0.054) | (0.013) | (0.081) | (0.019) | (0.017) | (0.014) | (0.065) | (0.083) | |
| R-squared | 0.001 | 0.003 | 0.007 | 0.003 | 0.003 | 0.001 | 0.009 | 0.004 | 0.009 | 0.013 | 0.008 | 0.003 | 0.004 | 0.037 | 0.004 | 0.009 |
| Block Fixed Effects | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Panel B: Balance between Control and Free-glasses (no education) group accounting for attrition in follow-up | ||||||||||||||||
| Free eyeglasses | −0.074 | −0.008 | −0.050 | 0.045 | 0.027 | 0.017 | −0.005 | 0.015 | −0.002 | 0.007 | −0.031 | −0.003 | 0.031 | 0.027 | 0.043 | −0.046 |
| (0.100) | (0.033) | (0.034) | (0.124) | (0.092) | (0.077) | (0.031) | (0.041) | (0.075) | (0.018) | −0.106 | (0.027) | (0.026) | (0.021) | (0.096) | (0.116) | |
| Attrition between baseline and evaluation | −0.170 | −0.164* | −0.081 | 0.158 | 0.143 | 0.015 | 0.139 | 0.115 | 0.211** | 0.023 | 0.062 | 0.083 | 0.116 | 0.181** | 0.058 | −0.278*** |
| (0.266) | (0.097) | (0.121) | (0.467) | (0.276) | (0.265) | (0.087) | (0.115) | (0.083) | (0.064) | −0.224 | (0.094) | (0.096) | (0.087) | (0.211) | (0.080) | |
| Interaction_attr2 | 0.277 | 0.272* | 0.077 | −0.452 | −0.332 | −0.120 | −0.200* | −0.261 | −0.019 | −0.057 | −0.101 | −0.056 | −0.088 | −0.039 | −0.150 | 0.088 |
| (0.428) | (0.158) | (0.173) | (0.547) | (0.348) | (0.378) | (0.106) | (0.182) | (0.159) | (0.077) | −0.353 | (0.123) | (0.127) | (0.128) | (0.287) | (0.122) | |
| Constant | 10.539*** | 0.614*** | 0.531*** | 2.392*** | 1.257*** | 1.135*** | 0.161*** | 0.464*** | 0.239*** | 0.622*** | 0.204** | 0.167*** | 0.084*** | 0.069*** | 2.192*** | 0.378*** |
| (0.064) | (0.022) | (0.021) | (0.079) | (0.057) | (0.046) | (0.022) | (0.029) | (0.054) | (0.013) | −0.081 | (0.019) | (0.017) | (0.014) | (0.065) | (0.083) | |
| R-squared | 0.002 | 0.003 | 0.003 | 0.001 | 0.001 | 0.000 | 0.004 | 0.003 | 0.008 | 0.001 | 0 | 0.001 | 0.005 | 0.014 | 0.001 | 0.011 |
| Block Fixed Effects | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
| Panel C: Balance between Control and Free-glasses (and education) group accounting for attrition in follow-up | ||||||||||||||||
| Education plus free eyeglasses | −0.103 | −0.024 | −0.032 | 0.155 | 0.039 | 0.116 | −0.001 | −0.010 | 0.010 | 0.007 | 0.176* | −0.002 | 0.045* | 0.029 | 0.004 | −0.013 |
| (0.119) | (0.032) | (0.030) | (0.122) | (0.070) | (0.079) | (0.030) | (0.043) | (0.076) | (0.017) | −0.101 | (0.025) | (0.025) | (0.020) | (0.106) | (0.118) | |
| Attrition between baseline and evaluation | −0.170 | −0.164* | −0.081 | 0.158 | 0.143 | 0.015 | 0.139 | 0.115 | 0.211** | 0.023 | 0.062 | 0.083 | 0.116 | 0.181** | 0.058 | −0.278*** |
| (0.265) | (0.097) | (0.121) | (0.467) | (0.276) | (0.265) | (0.087) | (0.115) | (0.083) | (0.064) | −0.224 | (0.094) | (0.096) | (0.087) | (0.211) | (0.080) | |
| Interaction_attr3 | 0.949** | 0.288** | 0.011 | −0.026 | −0.332 | 0.305 | −0.156 | −0.033 | −0.174 | 0.027 | −0.329 | −0.070 | −0.031 | −0.136 | −0.365 | 0.484*** |
| (0.370) | (0.131) | (0.151) | (0.591) | (0.327) | (0.374) | (0.108) | (0.140) | (0.122) | (0.081) | −0.292 | (0.119) | (0.123) | (0.112) | (0.256) | (0.120) | |
| Constant | 10.539*** | 0.614*** | 0.531*** | 2.392*** | 1.257*** | 1.135*** | 0.161*** | 0.464*** | 0.239*** | 0.622*** | 0.204** | 0.167*** | 0.084*** | 0.069*** | 2.192*** | 0.378*** |
| (0.064) | (0.022) | (0.021) | (0.079) | (0.057) | (0.046) | (0.022) | (0.029) | (0.054) | (0.013) | −0.081 | (0.019) | (0.017) | (0.014) | (0.065) | (0.083) | |
| R-squared | 0.014 | 0.004 | 0.002 | 0.003 | 0.001 | 0.006 | 0.003 | 0.002 | 0.004 | 0.002 | 0.008 | 0.001 | 0.009 | 0.009 | 0.004 | 0.010 |
| Block Fixed Effects | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES | YES |
Notes: Data source, baseline survey and evaluation survey; Study tested whether the effects of different treatments on vision knowledge and eyeglasses usage are biased due to sample attrition. For this, we first constructed a variable representing attrition (1 = attrition). Then, study used treatment variable, attrition variable, and interaction variable between the two to regress different baseline covariates and adjusted for school level cluster; robust standard errors in parentheses; ***significant at the 1% level; **significant at the 5% level; *significant at the 10% level.