| Summary: | 博士 === 國立嘉義大學 === 教育學系研究所 === 104 === Among the confirmatory factor analysis (CFA) of Structural Equation Modeling (SEM), the bifactor and second-order factor models are two alternative approaches to evaluate the constructs of intercorrelation among the factors assumed. This study compares and confirms the factor structure of Transformational Leadership Scale in Multifactor Leadership Questionnaire (MLQ) (5X-Short) and the factors’ predictions on external criterion variables. The study aims to find a better interpretation for Transformational Leadership Scale and thus to comprehend the relationships between the factors of transformational leadership and its variables. The study takes public elementary school adjunct section chiefs from the west of Taiwan as its research subject. A total of 650 official questionnaires are sent out, 539 valid questionnaires are responded and a response rate of 82.9% is achieved. Statistical approaches, including descriptive statistics as well as the bifactor and second-order factor models, are adopted to analyze the survey data and confirm the theoretical structure. Main research findings are as follows.
1.The bifactor model confirms that there is a necessity for all the five “specific factors” of Transformational Leadership Scale, including Idealized Influence (Attributes) (IA), Idealized Influence (Behaviors) (IB), Inspirational Motivation(IM), Intellectual Stimulation(IS) and Individualized Consideration(IC), to exist individually and independently from the “general factor”. This confirms the model of Transformational Leadership developed by Bass and Avolio (2004). By doing so, the scale’s items can be reduced.
2.The second-order factor model is in fact nested in the bifactor model. The former confirms that Transformational Leadership Scale has a factor structure of one “second-order factor” and five “first-order factors”, though a significant difference exists between the former and the latter. In addition, the second-order factor of the second-order model differs from the general factor of the bifactor model in terms of extracting the meaning of item variances.
3.After the factor structure of Transformational Leadership Scale is confirmed by the bifactor and second-order factor models, these two models fit the data well. It is also found that the bifactor model fits the data better than the second-order factor model, suggesting that the bifactor model provides more adequate representation of the scale. This shows not only the advantage of the bifactor model in terms of scale construction and evaluation but also the restriction of the second-order model being more strict.
4.Both the general factor and specific factors of Transformational Leadership Scale can predict most external criterion including work engagement, job burnout and turnover intention. Among all the prediction results, the specific factor of Individualized Consideration turns out to be the most significant, which implies that the transformational leadership behaviors of elementary school directors should take “Individualized Consideration” as their core value. However, the bifactor model allowed for easier interpretation of the relationship between the specific factors and external variables, over and above the general factor.
In the study, the researcher looks further into the implication of transformational leadership contained in the analysis findings and provides concrete suggestions for directors and section chiefs in elementary school as well as for future studies.
|
|---|
