Application of Regression Analysis to Identify Factors Affecting Crop Yield in Benue State

Agricultural productivity in Benue State, the North Central food basket of Nigeria, is influenced by multiple climatic, agronomic, and socioeconomic factors whose relative statistical contributions to crop yield variation have not been comprehensively quantified at the state level. This study applies multiple linear regression analysis to identify and quantify factors affecting crop yield among smallholder farmers in Benue State. Survey data were collected from 250 randomly selected farming households across Makurdi, Otukpo, and Gboko local government areas, capturing yield data for maize, yam, and cassava alongside predictor variables including rainfall, fertiliser application quantity, farm size, farmer education level, and access to extension services. Multicollinearity was assessed by variance inflation factors, and residual diagnostics confirmed model assumptions. The overall regression model explained 71 percent of variance in crop yield (R-squared = 0.71, F = 48.3, p < 0.001). Rainfall adequacy (beta = 0.38), fertiliser application (beta = 0.31), and extension service access (beta = 0.24) were the strongest statistically significant predictors. Farm size showed a significant positive effect, while farmer age was negatively associated with yield. Educational attainment was not independently significant after controlling for extension access. The study recommends that Benue State agricultural policy prioritise irrigation infrastructure, fertiliser subsidy programs, and expansion of extension officer coverage to maximise crop yield across the state. Keywords: regression analysis, crop yield, Benue State, agricultural statistics, smallholder farmers