Regression (Hierarchical) Analysis with a Moderator in SPSS 

Thank you. Glad it helped. Pls, like and share the video

Glad it was helpful!

Thank you very much

During a moderation analysis, if an interaction is not statistically significant, you may wish to remove the non-statistically significant results (after reporting the result), to see whether the results of the significant interaction term can increase. There is no strict rule on how to conduct a regression analysis with interaction terms. Ensure you specifically explain your reasons for every steps taken

Thank you. Every variable in your study should be part of your questionnaire.

Thank you sir. You do not need to mean-centre the dependent variable because, it is the one receiving the effects from the predictors (moderator and independent variables).

Thank you. The video is clear and clean. It may be internet issues. Thanks

All independent variables and moderator variables for Hierarchical Regression should be mean centred. For multiple regression analysis, do not mean centre them

Not recommended to run a regression with a data that is not normally distributed.

What about smart pls for not normal data

SPSS has provided you with the opportunity to perform the analysis, now you ask for a formula to manually calculate the analysis? The video is not about manual calculation.

Doc I was asking because I have seen in some papers they include formulas in there write up. Y= bo +Bx1+e, up to the end..

@@kalalakanyanda8418 The reason for using SPSS is to help you navigate the challenges associated with calculating the analysis by hand using the formula.

Thank you. Maybe, this video can help you ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-5v30M_I5eQM.html

Yes. It is possible. But ensure that you make a strong case as to why you wish to employ any statistical approach to arrive at your findings and conclusions. The most important thing is that the processes that led to your findings must be clear, concise and unambiguous.

You report your results. A non-significant interaction results will not have any interaction results.

@@Dr.Okolie if the moderator is significant but negative, what's your say on this Dr?

@@kalalakanyanda8418 Then, there is an interaction. Plot the interaction graph to learn the interaction slope. Normally, it may show a decrease effect on the dependent variable.

Also wanna know why to knock out inflex_x_friends (negatively significant), then knock out inflex_X_teachers (non-significant), but only keep inflex_x_parents, I guess this part was not well explained.

You will need to conduct a structural equation modelling or path analysis to achieve that.

To run a moderated mediation, you'll succeed with a path analysis or ProcessMacro rather than a multiple regression in SPSS