Methodological Evaluation of Quasi-Experimental Design in Assessing Yield Improvements at Water Treatment Facilities in Kenya

Water treatment facilities in Kenya have seen varying degrees of yield improvement over time, driven by different factors such as infrastructure upgrades and operational management practices. A quasi-experimental design was employed, where pre-treatment and post-treatment yield data from multiple facilities were analysed using regression discontinuity designs (RDD) with local polynomial smoothing techniques. Uncertainty around the estimated yield improvements is quantified through robust standard errors. The analysis revealed a statistically significant increase in water treatment yields by an average of 15% across all tested facilities, confirming the reliability and validity of QED methods for measuring such changes. Quasi-experimental designs provide a valuable tool for monitoring and evaluating yield improvements at water treatment facilities, offering insights into operational efficiency and system performance enhancements. Future studies should consider broader implementation of QED in various settings to validate the method's generalizability. Additionally, further research could explore the impact of different factors on yield changes. Quasi-Experimental Design, Regression Discontinuity Design, Water Treatment Facilities, Yield Improvement, Kenya The maintenance outcome was modelled as Yᵢₜ=β₀+β₁Xᵢₜ+uᵢ+εᵢₜ, with robustness checked using heteroskedasticity-consistent errors.