Quantifying the effects of one node on another node in path modelling is well-defined in univariate analyses, but is a more open problem when the nodes are multivariate. Attempts have been made before, though none are fully transparent and intuitive. We propose a new definition, which is motivated by simple orthogonalisations and then generalised to flexible regression.
Alternative definitions of effects in path models with multidimensional blocks / Næs, Tormod; Romano, Rosaria; Tomic, Oliver; Smilde, Age K.; Hovde Liland, Kristian. - (2025). ( 9th Scandinavian Symposium on Chemometrics (SSC2025) Keflavík, Iceland 15-18 giugno 2025).
Alternative definitions of effects in path models with multidimensional blocks
Rosaria Romano;
2025
Abstract
Quantifying the effects of one node on another node in path modelling is well-defined in univariate analyses, but is a more open problem when the nodes are multivariate. Attempts have been made before, though none are fully transparent and intuitive. We propose a new definition, which is motivated by simple orthogonalisations and then generalised to flexible regression.| File | Dimensione | Formato | |
|---|---|---|---|
|
Liland_et_al_SSC25.pdf
accesso aperto
Tipologia:
Versione Editoriale (PDF)
Licenza:
Dominio pubblico
Dimensione
771.71 kB
Formato
Adobe PDF
|
771.71 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
