EXPANSION OF ANALYSIS OF VARIANCE WITH SOME RANDOM FACTORS

• Main Author: Sintha Ully, Yuniar
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/55188
• Institution: Institut Teknologi Bandung

In an experiment, generally, the level of the observed factor is specifically selected by the experimenter to determine their effect on the associated response variable. However, in some conditions, the experimenter wants to know the effect of the possible set of levels on the observed factor so that the level is randomly selected. Cases like this could be solved with the Analysis of Variance (ANOVA) method with several adjustments. Furthermore, if the experiment has more than one response variable, a special method is needed which is the expansion of the ANOVA method, namely the Multivariate Analysis of Variance (MANOVA) method. In this undergraduate thesis, a case study using ANOVA and MANOVA methods is carried out on the intensity data of the gray scale of muscle and bone objects with frequency, current, and sequence as factors. By using the ANOVA method, the result shows that sequence always has a significant effect on muscle objects and that the interaction between frequency, current, and sequence always has a significant effect on both muscle and bone objects. In addition, it is also obtained that the factors and interactions that significantly influence the MANOVA method are a combination of significant factors and interactions in the ANOVA method for both muscle bone objects.