In the SEM-PLS 2.0 path analysis, my non of the control variable did not show any significant value on the dependent variable when I tested the mediation analysis (both for direct and indirect effect).
1) If I keep this control variable (insignificant) in the final model, will it make any sense?
2) Should I want to include the insignificant control variable for my analysis?
3) Is there any ways to make this insignificant to significant?
4)If I remove these control variable from my analysis will it make any problem?
Thanks a lot madam, If I know the path coefficient beta value and its standard error, can I find the p value? /
how can I find p-value by using path coefficient beta value and its standard error?
Morning,
Usually, you include a control only if there is a theory supporting the relationship. If you find some significant effects, you keep these results and report them. In this way you show that your theory holds even when controlling for other theoretically possible relationships.
If you find no significant effects, you can keep the controls. Your main effects will become, however, weaker, since (although insignificant) controls explain some of the variance in your dependent variables.
Therefore, researchers usually delete insignificant controls for "parsimonious reasons". It means that your hypothesised result is much stronger and other effects can be neglected.
Faithfully,
Eugene
Answer to your following question is:
You can calculate t-values/t-statistics by dividing beta value by standard error. The resultant value of t-statistics would tell you whether the beta is significant or otherwise. Any value of t above 1.96 would mean p-value less than 0.05, you can interpret your results likewise. t-value > 1.96 shows that the beta coefficient is significant.
Re: Thanks a lot madam, If I know the path coefficient beta value and its standard error, can I find the p value? /
how can I find p-value by using path coefficient beta value and its standard error?
Control variable in mediation analysis? - ResearchGate. Available from: https://www.researchgate.net/post/control_variable_in_mediation_analysis [accessed Mar 2, 2017].
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS