forum.smartpls.com

Hello all,

I am doing a research with SmartPLS3 and get a result like this:

Direct effect
Internet Quality -> Repurchase Intention : 0,128 (p-value:0,087) insignificant
Specific Indirect effect
Internet Quality -> Customer Satisfaction -> Repurchase Intention : 0,035 (p-value:0,205) insignificant
Total Effect
Internet Quality -> Repurchase Intention : 0,163 (p-value:0,026) significant

Can someone tell me what is the meaning of this results?
I hope someone can answer me ASAP
Thank you very much!

Significance is a strange concept that is also often criticized because of its dichotomous nature. Yes/no

In laymen terms (and not 100% correct): What you actually want to know is whether your effect is strong enough given the sampling uncertainty that you have to conclude that is is likely different from zero. The individual direct and indirect effect are small and the uncertainty around them is too large to make a definite conclusion (with 5% certainty). However, when you combine them the total effect is stronger while the uncertainty remains similar. Thus, you have more confidence in your total effect, then in the smaller direct (or even much smaller indirect) effect.

So in short: You probably have not enough power (sample size) to make a good conclusion about the weaker direct and indirect effects, while your sample size seem sufficient for the stronger total effect that is a combination of both.

BTW: This is a really good paper on why dichotomization into significant and not significant is not a good idea:
https://doi.org/10.1016/j.jbusres.2023.114086

Rigdon, E. E. (2023). How improper dichotomization and the misrepresentation of uncertainty undermine social science research. Journal of Business Research, 165, 114086.

The individual direct and indirect effect are small and the uncertainty around them is too large to make a definite conclusion (with 5% certainty). However, when you combine them the total effect is stronger while the uncertainty remains similar. Thus, you have more confidence in your total effect, then in the smaller direct (or even much smaller indirect) effect.

jmbecker wrote: ↑Thu Jul 20, 2023 10:32 am BTW: This is a really good paper on why dichotomization into significant and not significant is not a good idea:
https://doi.org/10.1016/j.jbusres.2023.114086 fnaf

Rigdon, E. E. (2023). How improper dichotomization and the misrepresentation of uncertainty undermine social science research. Journal of Business Research, 165, 114086.

yes, I believe it is a great analysis with multi-dimensional perspectives. Don't base your conclusions solely on whether an association or effect was found to be "statistically significant"