METHOD · mth_sem · v1
Structural Equation Modeling
结构方程模型
Represents relationships among observed and latent variables through measurement and structural models; causal interpretation requires substantive identification beyond fit.
方法身份
- 规范名称
- Structural Equation Modeling
- 别名
- SEM、latent variable modeling、path modeling
- 方法族
- latent variable / covariance structure modeling
以上名称和分类来自此版本资料卡,不代表方法之间存在已核验的关系或适用性。
适用问题
- latent constructs
- measurement models
- simultaneous path hypotheses
数据条件
- indicators with variation
- sample/information adequate for model complexity
- identified model
识别与模型假设
- model identification
- estimator/distribution assumptions
- measurement invariance when group/time comparisons are claimed
常见失败
- equating fit with causal truth
- indicator deletion only to improve fit
- ignoring nonidentification or inadmissible estimates
不能支持的结论
- causality from cross-sectional covariance alone
- proof that chosen model is uniquely true
诊断与核对
- identification
- residuals/parameter plausibility
- fit indices with caveats
- reliability/validity/invariance
可比较的替代方法
- multilevel models
- generalized regression
- PLS-SEM for different goals
软件与实现
- R lavaan
来源与参考
- lavaan: An R Package for Structural Equation Modeling
Yves Rosseel · 2012
打开原始来源 ↗
来源链接用于核对。来源可访问不代表已获全文或代码再分发授权。
对象关系
关系连接明确的对象版本。前置知识指学习或实现知识;搭配关系是编辑建议,尚未统计真实使用频率。
此对象版本尚未收录明确版本的关系。
关联研究主题
这些旧版关系没有绑定方法版本,只保留供回溯。当前版本的明确版本关系见“对象关系”;编辑建议不代表已验证适用性。
Digital Media and Communication →
Latent attitudes/path hypotheses
编辑关联 · 待领域复核Innovation and Entrepreneurship →
Latent capabilities/path hypotheses
编辑关联 · 待领域复核