This study aims to discuss the correlation of an increase in the bust with the width of the shoulder length and the thickness of the armhole as well as the influence of the three factors upon functionality of upper parts regarding the prototype design of new Bunka-style bodice. This study adopts a two-stage research method. First of all, a statistic analysis is performed on measurement of the upper bodies of 276 female students of Shu-Te University measured in 2005, and this study determines the correlation of the upper body of female university students at that time and the bust with the width of the shoulder length and the thickness of the armhole by using correlation analysis, principal component analysis, and cross tabulation. Secondly, a hypothesis of a clothes design for upper parts is made based on the bust, the width of the shoulder length, and the thickness of the armhole. Based on the bust range of 88±2cm~108±2 cm, there are a total of six groups with nine samples each, totaling 54 samples. Moreover, each group has three subjects, totaling 18 female university students aged 18~25, tested for functionality of upper parts, in order to evaluate satisfaction with comfortable and functionality of samples. Also, this study uses reliability analysis and MANOVA analysis to verify the drawing principles hypothesized by this study.
According to research findings, the stooping or hunchbacked body type is the most common of the all body types of female university students collected in 2005. Inferred from the result of correlation analysis, the bust has an influence on the thickness of the armhole with significantly high correlation of 0.805. Yet, the bust has low correlation with the width of the shoulder length, with a correlation coefficient of 0.103, representing no interrelationship between the bust and the width of the shoulder length. Moreover, inferred from evaluation results of satisfaction with comfortable and functionality, within the bust range of 88±2 cm, the best result falls in the width of the shoulder length of 11.5±0.5 cm and the thickness of the armhole of 11.5±0.5 cm; within the bust range of 92±2 cm~104±2 cm, the best result falls in the width of the shoulder length of 12.5±0.5 cm, and the thickness of the armhole will increase by 1 cm one time. Inferred from the foregoing results, indeed the bust has an influence upon the thickness of the armhole, but not the width of the shoulder length. In addition, the overall sensory test shows a confidence level of 83.4%, representing that the drawing for upper parts hypothesized by this study is applicable.