個別題項信度係指衡量變數用於測量潛在構面的一致性，即每個觀察變數的變異

個別題項信度係指衡量變數用於測量潛在構面的一致性，即每個觀察變數的變異量能被潛在構面所解釋的程度，也就是標準化因素負荷量(Standardized Loading)係數的平方，再進行檢測組合信度分析(composite reliability, CR)，學者Joreskog(1993)建議組合信度應該在0.6以上，以確定觀察變項內部一致性，分析結果發現某些題項對應構面因素負荷量(λ值)不足0.5，故予以刪除題項來表示。整體來說，本研究的衡量變數具有良好的表現。接著修正MI值(ModificationIndices, MI)來增加路徑，但未能表示該模式最終能進行研究的模式。M.I值的大小代表兩個變數存在高度相關，因此修正M.I提高模型解釋力外，應考慮到增加M.I值路徑邏輯上的合理性，循序漸進找到最佳模式進行研究，(吳明隆，2008)故此將刪除題項。本研究除了IN5、CO1、CO2、OR1、OR3、OR4、OR5、OR6、OR7、SA5依照M.I值數值最大依序刪除題項外，其餘題項M.I值皆屬平穩，修正M.I.值後將進行組合信度的檢測其結果顯示各個構面CR值都大於0.6這表示觀察變項內部均都達到一致性

Results (English) 1:

Reliability refers to the individual questions of measurement variables used to measure the potential dimensions of consistency, that is, each observed variable amount of variation can be explained by the potential dimensions of the degree, which is standardized factor loadings (standardized loading) coefficient squared , and then to detect composite reliability analysis (composite reliability, cr), scholar joreskog (1993) recommended composite reliability should be above 0.6 in order to determine the internal consistency of observed variables, the results found that certain questions of structure factors corresponding to the load quantity (λ value) of less than 0.5, it represented the questions be deleted. Overall, this study measured variables with good performance.Then fix mi values ​​(modificationindices, mi) to increase the path, but failed to indicate that the model will eventually be able to carry out the study mode. The size of mi values ​​represent two highly correlated variables exist, therefore amend mi improve the model explanatory power, shall take into account the increase in the value of the path mi logical rationality, step by step to find the best model to study (Wu Minglong, 2008) will therefore Remove the questions. In this study, in addition to in5, co1, co2, or1, or3, or4, or5, or6, or7, sa5 numerical value according mi largest order to delete the questions, the remaining questions of mi values ​​are among stable, fix miValue will be combined reliability test results show the various facets cr values ​​are greater than 0.6 which means that all observed variables have reached an internal consistency

Results (English) 2:

Individual item reliability refers to the measured variable is used to measure the potential dimensions of consistency, which means that each observed variable variable a surname can be explained by the potential dimensions of the extent of, that is, standardized load (Standardized Loading) coefficient squared, re-tested combination of reliability analysis (composite reliability, CR), scholar Joreskog (1993) Recommended combination of reliability should be 0.6 per cent, to determine the observation of internal consistency, the results found that certain questions correspond to dimensions of loading (lambda value) less than 0.5, the deleted item. Overall, the study measured variable has a good performance.Fix MI values (ModificationIndices, MI) to increase the path, but was unable to say which will eventually study mode. M.I value representative of the size of the two variables are highly correlated, the correction outside M.I increase explanatory power of the model, consideration should be given to increasing M.I value path logical rationality, step by step to find the best model for research, (Wu Minglong, 2008) as such topics will be deleted. Research in addition to the IN5, CO1, CO2, OR1, OR3, OR4 OR5, OR6 OR7, SA5 under M.I numerical maximum order value outside the deleted item, the remaining questions M.I values are stable, fixed M.I.Values will combine the reliability test results of individual dimensions of CR values are greater than 0.6, which indicates that the observation of internal consistency

Results (English) 3:

An individual topic credibility refers to the weight variable being used to survey the uniformity of latent construction scope, namely each observation variable changes the different quantity can by the degree that the latent construction scope institute explained that was also the standardized factor load (Standardized Loading) the square of coefficient, carried on the examination combination credibility analysis again (composite reliability, CR), the scholar Joreskog(1993) suggested to combine the credibility should over 0.6, by determining the observation variable interior uniformity, the results analysis discovered certain topic correspondence construction scope factor loads (Lamda value) less than 0.5, therefore gave to delete topic to indicate. The whole, the weight variable of this research has the good performance.Then revises the MI value (ModificationIndices, MI) increases the way, but cannot express that this pattern can carry on the pattern of research finally. The size of M.I value represents two variables to exist highly related, therefore revises M.I to enhance outside the model explanation strength, should consider that increases M.I value way rationality logically, proceeded in an orderly way to find the best pattern to study, (Wu Minglong, 2008), will therefore delete topic. This research deletes topic according to the M.I value value besides IN5, CO1, CO2, OR1, OR3, OR4, OR5, OR6, OR7 and SA5 most greatly in order, other topic M.I values all are steady, revises M.I.After the value, will carry on the combination credibility examination its result to show that each construction scope CR value is bigger than 0.6 this to express the observation variable interior achieves the uniformity