Title: Handling Assumption Violations In A Simple Regression Model
Authors: Hasnita
Volume: 10
Issue: 7
Pages: 256-259
Publication Date: 2026/07/28
Abstract:
Rice is the staple food in Indonesia, and the majority of the population relies on rice as their primary source of carbohydrates. Consequently, rice production is closely associated with population growth and demand. To examine the relationship between rice production and population, a simple linear regression model was developed. However, when the assumptions of linear regression are violated due to the presence of outliers or influential observations, the resulting model may produce biased and unreliable estimates. This study discusses several approaches to addressing assumption violations in simple linear regression. The methods employed include the Box-Cox transformation, logarithmic transformation, and the Weighted Least Squares (WLS) method to improve model performance and ensure more reliable regression results.