We have discussed in class Seva Mandir's program for (decentralized) flour iron supplementation where Hamlets (smaller than villages) were randomly allocated into this program (i.e., in Treatment and in Control). The treated Hamlets would get their flour fortified with iron as they take their grains to the Mills (the control Hamlets will not). When Seva Mandir tried to get information about hemoglobin measures (to test for anemia or iron deficiency) via blood tests there was an issue: Seva Mandir could not test for hemoglobin levels in the control group as much as in the treatment group. A large number of children and sick adults in the control group refused to take the test.

1- In your opinion should Seva Mandir be worried? If yes, explain what should Seva Mandir be worried about if not explain why? [hint: think of how the treatment group and the control group will change] Maximum length: 4 lines

2- Assume that you estimate the impact of the program and ignore this issue, would the estimate of the program's impact be accurate? If yes explain, if no explain what would happen to the estimated effect (i.e., underestimated or overestimated)? Maximum length: 2 line