Validity refers to the consistency

Validity refers to the consistency between the measured results and the measured content. Generally speaking, the higher the consistency, the higher the validity. The validity analysis is generally divided into content validity and structure validity. The content validity and questionnaire design measure whether the items contain all variables. Structural validity is to measure the rationality of the internal structure of the questionnaire which can be reflected by the items, and to study the structural validity by factor analysis.In this study, factor analysis is used to test whether the dimensions of online clothing customer experience, customer trust and customer satisfaction have reached the indivisibility. Factor analysis mainly uses two measurement indicators, kmo value and Bartlett's spherical test. Generally speaking, if the kmo value is greater than 0.8, it means the validity is high; if the value is between 0.7-0.8, it means the validity is relatively high. Good; if the kmo value is between 0.6-0.7, the validity is acceptable; in general, the kmo value = 0.6 is used as the limit of whether factor analysis can be performed on data. If the significance level of Bartlett value is 0.000, it is suitable for factor analysis. Through kmo value and Bartlett test scale, and principal component analysis by dimension reduction, it is generally considered that the correlation between a measurement item and its corresponding factor (i.e. factor load factor) is more than 0.5 is acceptable.According to table 5-9, five factors are extracted according to the standard with the eigenvalue greater than 1, and the extracted variance and the loaded cumulative variance are 78.147% ＞ 50%, meeting the standard. Therefore, according to the results of factor analysis, the dimensions of customer experience can be divided into five dimensions for research.In this study, the principal component factor analysis of 10 measurement items of online clothing customer trust is carried out, and the dimension of customer trust is divided by dimension reduction. The factor analysis results of customer trust are shown in table 5.11:According to table 5-11, three factors are extracted according to the standard with the eigenvalue greater than 1, and the extracted variance and the loaded cumulative variance are 71.127% ＞ 50%, meeting the standard. Therefore, according to the results of factor analysis, the dimensions of customer trust can be divided into three dimensions: integrity, ability and goodwill.Based on Schmitt's point of view, this study divides the online clothing customer experience into sensory experience, emotional experience, thinking experience, action experience and related experience when designing the questionnaire. Build Pearson correlation model to verify whether the five dimensions of customer experience are related to customer satisfaction.