Penerapan Regresi Linear Berganda untuk Menganalisis Faktor-Faktor yang Mempengaruhi Produksi Kopi di Kabupaten Dairi

Abstract

This study aims to apply a multiple linear regression model to analyze the factors that influence coffee production in Dairi Regency. Secondary data from 2023, including coffee production (tons/year), coffee cultivation area (hectares), number of coffee farming households (units), and elevation above sea level (masl), were obtained from the Central Statistics Agency (BPS) and the Agriculture Department in 15 sub-districts of Dairi Regency. The analysis was conducted using the Ordinary Least Squares (OLS) method with the assistance of SPSS software. The results indicate that the variable of coffee area (X1) has a positive and significant effect on coffee production, with a regression coefficient of 1.630 and a t-value of 14.7. Meanwhile, the variables of the number of coffee farming households (X2) and elevation (X3) did not show a significant partial effect. However, simultaneously, these three independent variables significantly influenced coffee production (F-value 415.94). The regression model explained 99.13% of the variation in coffee production (R² = 0.991). Based on classical assumption tests, the model residuals are normally distributed (Shapiro-Wilk p-value = 0.415), there is no multicollinearity (VIF < 10, tolerance > 0.10), autocorrelation (Durbin-Watson = 1.707), and heteroscedasticity (the Glejser test shows p-values > 0.05 for all independent variables). The conclusion of this study is that the multiple linear regression model constructed is valid, and the area of coffee plantations is the most significant factor influencing coffee production in Dairi Regency. This study is expected to contribute as a basis for decision-making by local governments and farmers in efforts to increase coffee production.