EFA before CFA & removing low outer loading

[kamellia.ch]

Hi. I am PhD student with the 300 samples thesis. I have two question appreciate if you guide me :

1- Is it necessary to apply EFA (Exploratory factor analysis) in SPSS before conducting your research in PLS and doing CFA (Confirmatory factor analysis)?

2- According to Hair et al., 2014, the outer loading 0.7 show high indicator reliability, and if between 4-7 , only you remove them when your composite reliability and AVE increase. after doing my algorithm, I needed to remove some lower outer-loading but remain some because the composite reliability didn't change. but at the other hand it means low indicator reliability and therefore you can not also claim you have convergent validity since reliability is necessary condition for validity. should I remove all the low outer-loading ? then the result will not be bias? (ps. the result have high composite reliability the only problem is indicator reliability)

appreciate your guidance

[cringle]

Usually, you can do the analysis directly in PLS. However, doing a CFA a priori may give you some orientation and certainly is not "harmful". Either way is fine; we usually directly do the analysis in PLS (the latent variables are different since they are established in the context of the model; these results differ from those of a CFA).

Best
CR

[kamellia.ch]

Thanks Dr.Ringle