The stepwise regression analysis method for estimating sorghum production in Karangmojo

Abstract

Abstract Sorghum is a multifunctional crop that can be utilized as a source of food, feed, and bioenergy. Sorghum is a plant that can adapt to land with optimal conditions, dry land with minimal nutrients and tolerant to pests and diseases. Sorghum has been widely cultivated in Indonesia, one in Gunungkidul, Yogyakarta. Sorghum production is influenced by several agronomic characteristics. The purpose of the study was to determine agronomic variables that affect sorghum production. The research was conducted in Karangmojo, Gunungkidul, Special Region of Yogyakarta, from October 2022 to March 2023. Sorghum varieties were local varieties as well as national superior varieties which are usually planted by local farmers. A total of 10 variables were analysed for their significance level on sorghum yield. The collected data were then processed using a multiple linear regression model (stepwise) using SPSS 16.0. The results showed that of the 10 agronomic variables observed, two variables contributed to the sorghum production, namely panicle weight and days to harvesting. The regression model from stepwise results is y = 8.884 + 0.036x3 – 0.062x10, with R2 = 0.754. This result indicates that the two independent variables are the main variables in determining sorghum production in Karangmojo.