RISE Working Paper 23/142 - July 2023 - Tackling Implementation Challenges with Information: Experimental

Second, they facilitated a re-allocation of tasks for the WEO, by circumventing the role of DEOs in (i) reading the reports, (ii) redirecting WEO efforts based on the findings of the report, and (iii) following up on WEO efforts related to the findings of the WSV report. [...] 11 treatment arms and 22 percent of schools in the control received a WSV by the start of midline data collection; and virtually all schools in our treatment groups and 43 percent of schools in our control group had received a WSV by endline (Panel A, Table 2) As a result of the staggered roll-out of the WSVs, there is variation in elapsed time between the date of the WSV and date of data collecti. [...] The survey included questions about the frequency and length of school visits, the activities they typically perform when visiting schools, their exposure to the new WSV framework, recollection of the main recommendations that were made in the WSV report, and the actions that they took in response to the report. [...] To minimize the risk of over-rejection of the null hypothesis due to multiple comparisons, we created indices of the main outcomes, by taking the mean of the standardized score of all the indicators relating to the same outcome (Kling et al., 2007). [...] The attriters in the Visit&Text arm are slightly older (0.37 years) than the attriters in the control, but the coefficients on the treatment dummies show that the sample remains balanced in terms of baseline characteristics, when restricting the sample to non-attriters.