How to model formative latent variable in R using lavaan, and also add control variables -
i want structural equation model analysis.
all 3 concepts formative, means these measurement variables construct latent structure. not reflective.
and want control demographic variables.
the analysis use r , learnt lavaan in statistics class, curious whether can give more information on want do.
thanks ahead.
in order used formative constructs need use partial least square sem (pls-sem) , not covariance based sem (cov-sem).
the underlying assumption of 2 technique differs, pls assumes because constructs formative there no error in measurement while cov-sem, has defaults assumption of reflexive measurement , takes account error terms while analyzing model
lavaan package cov-sem one, can try plspm using formative items
here there guide on how use plspm pakage: http://gastonsanchez.com/pls_path_modeling_with_r.pdf
Comments
Post a Comment