

RDSEM longitudinal mediation 

Message/Author 


If I want to test whether there is a nonzero indirect pathway in the crosslagged relationships from x > m > y, would the following syntax be correct? MODEL: %WITHIN% x m y ON TIME; m_to_m  m^ ON m^1; x_to_x  x^ ON x^1; y_to_y  y^ ON y^1; x_to_m  m^ ON x^1; y_to_m  m^ ON y^1; m_to_x  x^ ON m^1; y_to_x  x^ ON y^1; x_to_y  y^ ON x^1; m_to_y  y^ ON m^1; %BETWEEN% m x y; m_to_mm_to_y; [x_to_m m_to_y] (A B); MODEL CONSTRAINT: NEW(IND); IND = A * B; Thanks! 


The syntax looks ok. But the topic of mediation with longitudinal data is complex. See for instance the MaxwellCole articles  2007 in Psych Methods and 2011 in MBR. 


Excellent. Thanks Bengt, I'll have a look at those papers now 

Back to top 

