英国萨塞克斯论文代写:定量研究方法

Keywords:英国萨塞克斯论文代写:定量研究方法

本研究采用了Creswell(2003)所描述的定量方法，强调利用定量调查来确定泰国南部三个省份的伊斯兰私立中学教师认为的变量、战略领导和变革领导风格以及教师承诺之间是否存在影响。结构方程建模(SEM)用于确定假设效应模型在多大程度上得到支持，以及假设概念模型在多大程度上适合相关数据。结构方程建模(SEM)是首选的方法，因为许多先前的研究都支持SEM在这类研究中的应用(参见Clegg et al. 1997, Neilson 1997)。SEM也被选择，因为它能够定义和测试一种全面的“系统应急方法”(Hiltz, 1994)类型的理论模型。例如，Chin (1998, vii)提到，“在应用正确的时候，基于语义的程序相比于第一代技术，如主成分分析、因素分析、判别分析、或多元回归，都具有显著的优势，因为研究者在理论和数据之间的相互作用具有更大的灵活性”。与这些类型分析中经常使用的“第一代”技术相比，SEM的一些优点包括:(1)估计变量的直接、间接和总影响;(2)定义和研究潜在构念之间的关系;(3)估计模型中其他变量对每个潜在构念的方差;和(4)估计与每个观测和潜在变量相关的误差项。
英国萨塞克斯论文代写:定量研究方法
This research adopted a quantitative approach as described by Creswell (2003), by emphasizing the utilization of quantitative surveys to determine if the effects existed between the variables, strategic leadership and transformational leadership styles, and teacher commitment as perceived by Islamic private secondary school teachers in three provinces in Southern Thailand. Structural Equation Modeling (SEM) is used to determine to what extent the model of hypothesized effects is supported, and how well a hypothesized conceptual model fits the associated data.The Structural Equation Modeling (SEM) is preferred because many previous studies supported the employment of SEM in this kind of research (see e.g. Clegg et al. 1997, Neilson 1997). SEM is also selected because of its ability to define and test a comprehensive "System Contingency Approach" (Hiltz, 1994) type of theoretical models. For instance Chin (1998, vii) has mentioned that, "when applied correctly, SEM-based procedures have substantial advantages over first-generation techniques such as principal component analysis, factors analysis, discriminant analysis, or multiple regression because of the greater flexibility that researcher has for the interplay between theory and data". Compared to these "first generation" techniques often used in these types of analysis, some of the advantages of the SEM include the ability to: (1) estimate the direct, indirect, and total effects of variables; (2) define and investigate relationships among latent constructs; (3) estimate the variance accounted for in each latent construct by other variables in the model; and (4) estimate error terms associated with each observed and latent variable