Page 1 of 1

IPMA - Include LVs with non-significant total effect ?

Posted: Sun Aug 11, 2019 10:56 pm
by agalvez
Hi

When asessing structural model relationships, what happen if some LVs’ total effects (importance) are non-significant explaining the target LV? Do I have to include them in the IPMA? Does this make sense?

Another issue... Negative total effects are included in the map as "negative" or as "absolute values" ? I think I have to include them as absolute values, since the purpose is to compare their relative importance with other constructs. Is this correct?

Thank you in advance

Re: IPMA - Include LVs with non-significant total effect ?

Posted: Wed Aug 21, 2019 4:05 pm
by cringle
Insignficant total effects and negative ones (when you expect positive ones) are not reasonable for running an IPMA.

Re: IPMA - Include LVs with non-significant total effect ?

Posted: Thu Aug 22, 2019 12:37 pm
by agalvez
Thank you Prof. Ringle.

Then, insignificant total effects are excluded from the IPMA analysis, focusing on the significant ones.

Best regards.