There is mounting evidence suggesting that there is considerable heterogeneity in student preparation in primary schools in India. The largest education non-profit in India, Pratham, developed a pedagogical approach called "Teaching at the Right Level" (TaRL) that provides differentiated instruction to primary schoolchildren to address heterogeneity in preparation. However, persuading teachers to change business-as-usual instruction to incorporate this--or any other--method of differentiated instruction is challenging. Therefore, it seems worth understanding whether making the results of the assessments of basic reading and arithmetic skills more salient to teachers, and signaling the government’s commitment to improving these results, would encourage teachers to adopt TaRL on their own during school hours. We plan to randomly assign 200 public primary schools in the Indian state of Karnataka to: (a) a "control" group that would implement TaRL and receive no additional interventions; or (b) a "treatment" group that would also implement TaRL and receive periodic school report cards with results from student assessments of basic reading and arithmetic skills, accompanied by a letter from a high-level government official communicating the importance of improving those results. The study would focus on schoolchildren in grades 4 and 5, where TaRL has been found to be most effective.
External Link(s)
Citation
Banerjee, Abhijit, Alejandro Ganimian and Shobhini Mukerji. 2017. "The impact of diagnostic feedback on differentiated instruction: Experimental evidence from India." AEA RCT Registry. August 15. https://doi.org/10.1257/rct.2161-1.0.
Former Citation
Banerjee, Abhijit, Alejandro Ganimian and Shobhini Mukerji. 2017. "The impact of diagnostic feedback on differentiated instruction: Experimental evidence from India." AEA RCT Registry. August 15. https://www.socialscienceregistry.org/trials/2161/history/20441.
The feedback component would include: (a) a brief and user-friendly report on the basic math and reading skills of students in grades 4 and 5, based on the teacher-administered tests; and (b) an accompanying letter from the district magistrate highlighting the importance of improving children’s performance on these assessments by implementing OK. There will be two reports: one shortly after the baseline and another one shortly after the midline.
Intervention Start Date
2017-12-01
Intervention End Date
2018-03-31
Primary Outcomes (end points)
The study would evaluate the impact of the feedback intervention on the learning outcomes of school children, as measured by both the ASER tests and independent written assessments in basic reading and arithmetic skills, as well as the implementation of TaRL in the classrooms.
Primary Outcomes (explanation)
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
The 200 schools selected for the study would be randomly assigned to one of two groups: (a) “control group” that would implement TaRL, but not receive feedback (~100 schools); or (b) a “treatment group” that would implement TaRL and receive feedback (~100 schools).
Experimental Design Details
Randomization Method
The randomization (that is, identifying schools as treatment or control schools) will be done by running a Stata code on a computer in office.
Randomization Unit
The randomization will be done at the school level. The sample of schools will be stratified by district and size (i.e., number of students enrolled in grades 4 and 5).
Was the treatment clustered?
Yes
Sample size: planned number of clusters
The study will consist of 200 schools.
Sample size: planned number of observations
There will be 10,200 observations consisting of:
- 600 teachers (2 teachers and 1 principal per school) and,
- 10000 students (an average of 50 students per school)
Sample size (or number of clusters) by treatment arms
A treatment group of about 100 schools will implement TaRL and receive feedback.
A control group of about 100 schools will implement TaRL, but not receive feedback.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)