Law and economics research methods new thinking - Introduction to Structural Equation Modeling
Abstract: Structural equation modeling in the social sciences are more and more attention and application, as compared with traditional research tools, it has obvious advantages. Theory of the firm to Law and Economics of Executive Compensation study, for example, may well reflect the relationship between various factors and impact of the path,nike sale, which for the Economic Analysis of Law in the corporate governance theory provides a more clear structural relationship.
Keywords: Law and Economics; structural equation model; path analysis
Abstract: SEM has been given more thought than ever in the researches of social science as it has some unipue advantages compared with the traditional methods. With regard to the compensation for manager from the perspective of corporate theory in law economics, the paper expounds the intervelationship of those elements involved that clearly represent the structural relationship of corporate governance.
Key words: law and economics; SEM; route analysis
structural equation modeling (Structural Equation Model, referred to as the SEM) as a types of multivariate statistical techniques, have quickly received widespread application. In the early 20th century, 70 some scholars (Joreskog, 1973; Wiley, 1973) to factor analysis, path analysis, integration of statistical methods, structural equation model presented preliminary concepts. Then Joreskog and its partners to further develop the matrix model analysis techniques to deal with the analysis of covariance structure is proposed measurement model and structural model of the concept led to the development of SEM. Structural equation model for the fact that a verification of one or more independent variables in one or more of the dependent variable, a set of interrelationships between the multivariate analysis program, in which independent and dependent variables can be continuous, but also discrete. In addition, academic activities, according to Hershberger (2003) study from 1994 to 2001 between the literature found that by 2003, both in published structural equation model of the relevant paper journals the number of journal articles, the number of structural equation models extended from the multivariate analysis and other areas, have significant growth, structural equation modeling show is already a mature and highly valued the knowledge and technology. In addition to its unique structural equation model
Covers a wide range of structural equation modeling of the original multivariate data analysis methods for sequencing, will be given class and spacer and the ratio of scale in management, economics and other social sciences and natural sciences statistical evidence study gradually a large number of applications. Integration of structural equation modeling path analysis, confirmatory factor analysis and general statistical test method that analyzes the mutual causal relationship between variables,
Ray-Ban RB4141, including factor analysis and path analysis of the advantages. At the same time, it makes up for the shortcomings of factor analysis, taking into account the error factors, path analysis does not require the assumption by the restrictions. Structural equation model can be also analyzed the relationship between a group of formula,
ray ban 8301,oakley
sunglasses, especially the causal equation. This can handle the relationship between multiple variables the ability to help researchers conduct exploratory analysis and confirmatory analysis. When the theoretical foundation is weak, the relationship between multiple variables is not clear and could not be confirmed when the relationship between factors, exploratory analysis can be used to analyze the relationship between variables; when the study has theoretical support when the analysis can be applied to verify to verify the existence of the relationship between variables.
Although the structural equation model in a number of disciplines has been widely used, but few researchers on structural equation modeling process to make that many researchers have access to relevant information will be difficulties. Thus, by simply describes how the rational application of law and economics in the structure equation, structural equation modeling process are discussed, I hope the role of better ideas. In addition to the management and economics, the structural equation model is particularly useful, it provides a systematic way of thinking, in addition to doing research a specific hidden variables, but also for the qualitative change to quantitative research opens up new ideas.
First, the basic principles of structural equation modeling structural equation model assumes
a potential causal relationship between variables and potential variables can be significant variables were expressed by a group, is one of several significant linear combination of variables. Verified significant covariance between variables can be estimated the coefficient of linear regression model to test the statistical model assumed in the course of the study is appropriate, if it is proved that the model assumed the right to say the potential variables the relationship between reasonable.
(A) General equation model
1.
With a latent variable structural equation structural equation model has two basic modes: structural and measurement mode (HAU Kit-tai, Wen Zhonglin, into sub-Juan, 2004). Exogenous variable structure model shows the potential and the potential endogenous causal relationship between variables, this relationship graphically expressed in the form known as the road map.
Structural equation η = Bη + Γξ + ζ
equation for the ξ potential exogenous variables (latent variables) matrix; η as a potential endogenous variable (the dependent variable matrix potential;) for the structure coefficient matrix, Γ structural model that represents a potential independent variable matrix ξ η of the potential impact of the dependent variable matrix; B for the structure coefficient matrix, it represents a potential structural model is the dependent variable matrix η is the interaction between the components; ζ structural equation residuals matrix.
2.
Measurement mode measurement equation description of the potential variables η, ξ and measurement variables y, x relationship. Measurement equation can be expressed as Y = ΛYη + ε,
ray ban rb3025 silver mirror, where Y is the measurement variable η matrix; ΛY to measure the coefficient matrix, which indicates a potential endogenous variable (latent dependent variable) matrix η and the relationship between measured variables Y; η as a potential endogenous variable (latent dependent variable) matrix; ε equation for measuring the residual matrix.
Measurement equation can be expressed as X = ΛXξ + δ, the equation X variable matrix for the measurement of ξ; ΛX to measure the coefficient matrix, which indicates a potential exogenous variables (latent variables) matrix η and its The relationship between measured variables X; ξ as a potential exogenous variables (latent variables) matrix; δ for the measurement equation of the residual matrix.
(B) of the model estimation and evaluation
model once set, and then to be based on observed variables for the variance and covariance estimation. Structural equation model estimation process is different from traditional statistical methods, it is obtained from a sample of the phenotypic covariance matrix of variables, or correlation matrices deduced an implied variance and covariance matrix, or correlation matrix, the matrix each element as close as possible the sample variance of observed variables in the covariance matrix of the corresponding element. If the model is set correctly, will be very close to the variance-covariance matrix, so the estimation is made using a special fitting function derived variance-covariance matrix of the phenotypic covariance matrix of variables the difference between low as possible.
Estimate the model has been developed from a large number of estimation methods such as maximum likelihood estimation, generalized least squares, weighted least squares and non-progressive interference-free weighted least squares method. Is more commonly used maximum likelihood estimation, generalized least squares, but different estimation methods have their advantages and disadvantages. The main purpose is to discuss general principles of structural equation model, the relevant calculation method can refer to the relevant information.
Evaluation of the model, involves the degree of model fit the data. The overall model fit on many levels and standards of measure, the most commonly used fit index is the goodness of fit chi-square test (χ2), the chi-square value of the fitting function value can be used directly derived, equivalent to fitting function value and the sample size minus 1 of the product. Chi-square value on the size and sample size, so they have been since the development of goodness of fit index (GFI), the modified goodness of fit index (AGFI), the absolute goodness of fit index, goodness of fit index value, the provincial thrifty goodness of fit index, from the goodness of fit index, and the square root square mean residual (RMR), Benteler - Bo Neite specification fit index (NFI), the square root error of approximation (RMSEA) and information standards index. According to the characteristics used to verify the data, sample size and select the appropriate assumptions evaluation.
Second, structural equation modeling
the main features of the basic principles of structural equation model from the introduction shows,Tory Burch outlet, the structure equation with authentication features. The researchers used certain statistical tools to deal with the complex theoretical model,
ray ban occhiali, based on the consistency between models and data level, to make a proper evaluation of theoretical models, which assume in advance to confirm or falsify theoretical models. Process can be summarized from the following features.
(A) of the theoretical a priori
More articles related to topics： abercrombie and fitch First, the reality of optimal tax theory and personal income tax system to imp Tory Burch Outlet Misty mountain _ Pakistan Mbt Schuhe Günstig Commerce Department the first
ray ban 8301ray ban occhialiRay-Ban RB4141
Hybrid Mode
