Use of the Qq Plot Method in Determining Normality of Regression Data Apart from Using the Kolmogorov Smirnov and Shapiro Wilk Methods

Abstract

This research investigates the utilization of the QQ Plot method in evaluating the normality of regression data, while also comparing it with conventional statistical tests such as Kolmogorov-Smirnov and Shapiro-Wilk. The normality of data is a crucial aspect in regression analysis, and the selection of effective testing methods can significantly impact the validity of the analysis results. In this study, a holistic approach is taken by combining statistical and visual analysis through the QQ Plot method. Normality tests are conducted on the unstandardized residuals of the regression model. The results of the Kolmogorov-Smirnov and Shapiro-Wilk statistical tests indicate significant non- normality in the data. Furthermore, the QQ Plot graph is employed to provide a deeper visual understanding of how closely the data distribution approximates a normal distribution. The research findings highlight that the QQ Plot method can offer a clearer perspective on the distribution of regression data, especially in detecting patterns that may be challenging to identify with traditional statistical tests. While statistical tests provide necessary quantitative evidence, QQ Plot can be a valuable tool in interpreting the characteristics of the distribution more profoundly. In conclusion, understanding the normality of regression data can be enriched by incorporating the QQ Plot method alongside conventional statistical tests. The combination of these two approaches can provide a more comprehensive perspective, assisting researchers in making more informed decisions in regression analysis. This research lays the foundation for further exploration of the use of QQ Plot in the context of more complex data analysis.