Linear model implies a covariance matrix matrix
The video explains the application of path analysis tracing rules for calculating covariances in statistical models. It highlights the extension of these rules, originally used for correlations, to also encompass covariances. The significance of two-headed arrows in these rules is emphasized, indicating their role in quantifying correlations, covariances, and variances. The video demonstrates how to make variances explicit in a model by adding variances of specific variables and an error term, which is essential for effective tracing.
Path analysis tracing rules are applied to both simple recursive models and more complex multiple equation models, like mediation models. The video discusses the challenges in estimating variances in complex models, especially when dealing with latent variables. It shows how to calculate the covariance between different variables, starting from one variable and tracing back to a two-headed arrow before moving to another variable. The video also touches on the concept of reduced form equations in econometrics, explaining how endogenous variables in a path diagram can be expressed as linear functions of exogenous variables. Finally, the importance of ensuring that each path in a model has a single variance or covariance for accurate calculations is underscored.
Slides: https://osf.io/dtyg5