Analisis Kepuasan Menggunakan Regresi Logistik Ordinal di Sonatine Piano Pematang Siantar
Satisfaction Analysis Using Ordinal Logistic Regression at Sonatine Piano Pematang Siantar
Abstract
Ordinal logistic regression is used to analyze the relationship between ordinal scaled dependent variables and independent variables. In this study, ordinal logistic regression is applied to determine the factors that influence student satisfaction at Sonatine Piano Academy. Music, especially piano, plays an important role in life, and non-formal learning such as piano lessons is one of the methods to support the learning process. However, in recent months, there has been a decline in the number of students at Sonatine Piano Academy, which affects the reputation and sustainability of the business. To reduce student attrition, it is necessary to analyze the factors that influence student satisfaction as the level of satisfaction affects the sustainability of the course at Sonatine Piano Academy. This study uses four independent variables, namely teaching aspects, understanding aspects, environmental aspects, and facility aspects with the dependent variable being the satisfaction of Sonatine Piano Academy students. The method used in this study is ordinal logistic regression with 30 respondents. Of the 30 total respondents, then look for any variables that have a significant effect on the dependent variable. The results showed that the test of model fit or goodness of fit of the Deviance method was 24,289 with the value of X(0,05,23)2 was 37,652. Because 24,289 > 37,652, it can be concluded that the ordinal logistic regression model is declared suitable. Variables that significantly affect student satisfaction at Sonatine Piano are obtained from three aspects of the four aspects studied, namely the teaching aspect, the understanding aspect, and the facility aspect. Overall, the teaching aspect, understanding aspect, environmental aspect, and facility aspect affect student satisfaction by 87.64%, while the remaining 12.36% is influenced by other variables from the model.
Collections
- Undergraduate Theses [1407]