Abstract
This study examined the relationship between observed demand, executed production, and defective units in order to provide useful evidence for supply and production optimization. A quantitative, explanatory, and applied approach was adopted under a non-experimental, retrospective longitudinal design using monthly records from January 2018 to December 2024. Rechazo_Total was defined as the dependent variable, while Ventas and Produccion_Total were used as explanatory variables. A multiple linear regression model estimated by OLS was implemented with an 80/20 temporal validation split. The findings showed that Produccion_Total maintained a positive, robust, and statistically significant effect on total rejections, whereas Ventas lost significance once production was controlled for. Although the model achieved relevant explanatory capacity, its out-of-sample predictive performance remained limited, with evidence of heteroscedasticity and autocorrelation. It is concluded that defect reduction depends mainly on interventions focused on production intensity and the operational pressures generated within the manufacturing process.