The comparison of goodness index of structural equation model.
Changyi Wang, Dapeng Wang, Xiaowen Zhao
Abstract
Changyi Wang, Dapeng Wang, Xiaowen Zhao
Abstract
Structural equation model (SEM),which is a method of building,estamating and testing causality modeling,has been being used widely in the fields of psychology,sociology,behavior science and medicine. Goodness index is usually used for evaluating the goodness effect of models,but there are more than 40 indexs for SEM,and the results of different ones are not same. So which indexes are credible becomes a concerned topic for users. On the base of collecting many literatures,we summarized systematically the category and characters of indexes of SEM. By the method of comparing analysis,we find NNFI and RMSEA are credible,the second is CFI,RNI and Mc. When estamating parameters with ML or GLS,we find SRMR has better effect.
OpenAlex reports 9 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Structural equation model (SEM),which is a method of building,estamating and testing causality modeling,has been being used widely in the fields of psychology,sociology,behavior science and medicine. Goodness index is usually used for evaluating the goodness effect of models,but there are more than 40 indexs for SEM,and the results of different ones are not same. So which indexes are credible becomes a concerned topic for users. On the base of collecting many literatures,we summarized systematically the category and characters of indexes of SEM. By the method of comparing analysis,we find NNFI and RMSEA are credible,the second is CFI,RNI and Mc. When estamating parameters with ML or GLS,we find SRMR has better effect.
Key concepts: Structural equation modeling, Goodness of fit, Index (typography), Econometrics, Statistics, Mathematics, Causality (physics), Psychology