Hi
I need some help for testing the mediation in a single Model--------
Model is bit complex--------------and it involves 2 mediators
I have 6 reflective variables---- the reliability and validity of these variables have been established--------------
I1 -> I2
I2 -> M1
I2 -> M2
M1 -> D1
M1 -> D2
M1 -> D3
M2 -> D1
M2 -> D2
M2 -> D2
M1 & M2 Mediates the relationship between I2 and D1, D2 & D3
All paths (bootsrap 500) are significant (except one i.e., M1 -> D2)
I tested the mediation by employing the procedure outlined in Preacher, K. J., & Hayes, A. F. (in press). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods.
1. Specific indirect effects of 6 paths were calculated i.e.
I2 -> M1 -> D1
I2 -> M1 -> D2
I2 -> M1 -> D3
I2 -> M2 -> D1
I2 -> M2 -> D2
I2 -> M2 -> D3
2. The significance of each specific indirect effects was tested using the procedure outlined by Preacher, K. J., & Hayes, A. F. (in press).
For my data all indirect effects were insignificant at 0.05 level but were significant at 0.10 level.
a) Is this a correct approach to be applied to PLS?
b) Can I interpret results at significane level of <0.10 significant
Regards
Jahanvash karim