Table 1.

Systematic Review Inclusion Criteria.

Criteria	Included studies	Excluded studies
Population	General program participants in L&MICs, where benchmark and NRS replication study draw on participants from the same target population and time period.	Between-study comparisons with no overlap in treatment group samples for causal benchmark and comparison (e.g., Glewwe et al., 2004).
Intervention and comparator	Any social or economic development intervention and any comparison condition (e.g., no intervention, wait-list, and alternate intervention).	Clinical or bio-medical interventions, or interventions conducted among populations in high-income country contexts (e.g., Fretheim et al., 2013)
Benchmark study design	Within-study comparisons reporting results of a benchmark randomised study, where randomisation was done by researchers or administratively.	Within-study comparisons where the causal benchmark did not use randomised assignment (e.g., Friedman et al., 2016).
NRS study design	Within-study comparisons reporting results of NRS comparison replication using any method (e.g., regression analysis applied to cross-section or panel data (DD), IV estimation, statistical matching, and RDD) from same target population and using the same outcome as benchmark study.	Within-study comparisons where target population differs from benchmark, for example, due to mismatch between ATE and LATE (Urquieta et al., 2009; Lamadrid-Figueroa et al., 2010).