提問:
蕭教授:
您好,最近在學習PLS-SEM有關smartpls軟體。 看了您編著的《統計分析入門應用:SPSS分析+SmartPLS3》第三版一書受益匪淺,有以下疑問向您請教:
Q1:反應性指標是否也需要共線性檢測? (國內許多實證論文中,反應性指標並未對共線性進行檢測)
Ans: If your correlations are high, it is better to have VIF report (共線性偵測)
Q2:關於路徑效果報告中,間接效果是否一定要在論文中報告?
Ans: It is up to your research.
If you want to discuss total effects, you must report direct, indirect (間接效應), and total effects.
Further, if you want to compare effects of factors, it is better also to report indirect (間接效應).
Q3:關於模型解釋力及模型報告中,我看到很多國內的文獻,僅報告了R2,是否一定需 要報告f2、q2呢?
Ans:
You must report f square under 1) moderation effect and 2) small path coefficient but significant.
You must report Q square for prediction.
Thanks for your questions and good luck to your research.
Prof. Mac Shiau (Wen-Lung Shiau)
請先 登入 以發表留言。