Causal Impact Studies Methods

SPIA commissions and/or manages a portfolio of impact evaluations, each designed to match the strongest feasible method to the research question and context. These studies are commissioned through an open, competitive, peer-reviewed process involving external researchers and CGIAR centers. The methodological toolkit for causal studies typically includes:
 

Randomized Controlled Trials (RCTs)

Where feasible, random assignment of an intervention creates the cleanest comparison between treatment and control groups. SPIA supports RCTs when innovations are being introduced in a phased or experimental manner that permits randomization.
 

Quasi-Experimental Methods

Many CGIAR innovations cannot be randomized. In such cases, SPIA employs rigorous quasi-experimental designs including:

• Difference-in-differences: Comparing changes over time between adopters and non-adopters to control for pre-existing differences
• Instrumental variables: Using external variation that may influence adoption but not outcomes directly, to isolate causal effects
• Regression discontinuity: Exploiting eligibility thresholds or geographic boundaries to identify impacts
• Matching methods: Constructing statistical comparison groups based on observable characteristics
   

Mixed Methods

Quantitative estimates of impact are strengthened when paired with qualitative evidence that explains how and why impacts occur or fail to. SPIA increasingly integrates qualitative approaches, especially when studying questions relating to policy impacts of CGIAR, to capture the mechanisms, contextual factors, incentives, and unintended consequences that quantitative data alone may miss.

Similarly, spatial methods may also be integrated into causal impact studies where relevant, depending on the research question and country context.

The causal impact studies (previously and currently) commissioned by SPIA are presented alongside reach evidence on the respective Country Studies, where findings from both work pillars are synthesized to provide a comprehensive picture of CGIAR's contributions within each country.

Building Capacity Across CGIAR

Generating rigorous causal evidence requires skilled researchers and institutional support, not just within SPIA, but across CGIAR as a whole. Per its mandate, SPIA plays an active role in strengthening impact assessment capacity system-wide through several channels:

• MELIAF and Impact Assessment Focal Points: SPIA works closely with the Monitoring, Evaluation, Learning, and Impact Assessment Function (MELIAF) grant, the Impact Assessment Community of Practice (IA-COP) and with SPIA’s own Impact Assessment Focal Points embedded across CGIAR centers. This includes convening focal points, providing inputs within MELIAF for CGIAR-wide decisions relating to impact assessments, and building a networked community of practice.
• Matchmaking and study design support: SPIA helps connect CGIAR impact assessment researchers with centers with external academics, facilitating timely and mutually beneficial partnerships. This matchmaking role lowers the barrier for CGIAR researchers to initiate high-quality evaluations of their own work.
• Training and guidance: Based on ongoing research and stakeholder demands, SPIA periodically publishes technical guidance notes and delivers trainings/workshops on causal inference methods, study design, and evidence standards tailored to CGIAR’s context. These aim to equip impact assessment researchers within CGIAR centers to both design studies and critically interpret impact evidence.

Through these efforts, SPIA aims to build a culture where rigorous impact evaluation is not confined to SPIA's own portfolio but becomes a routine part of how CGIAR assesses and improves its research investments.