To check the moderation effect on indirect effect, I chose high and low levels of my moderator in a hypothesis. It assumes that indirect effect gets improved under low-level moderation, and lessens at high-level moderation.
Now, the effect is high at low-level of moderation, and low effect at high-level of moderation (see given values below). However, statistical results are different when it comes to interaction between (MED x MOD). For instance, results illustrate negatively significant interaction between MED and MOD. Interaction graphs of MED and MOD are as per expectations, but, below mentioned values at moderation levels are all positive.
Therefore, (a) how to identify whether shown values at moderation levels are also negatively significant (no subtraction sign), OR those have significant and positive effect at all levels?
(b) Does moderation has negative effect at all level, and it always shows values positively, or moderation is actually positive in my case, thus it is showing it positive values at the three levels? (c) And if all values are positive/negative at levels, how to interpret and write them to support or reject given hypothesis (see sentence no. 02)
Values at three levels of moderation are as follows;
Low = .45 Effect p< 0.00
Mean = .23 Effect p< 0.01
High = .16 Effect p< 0.03
Interaction;
MEDxMOD = B -.540 p< 0.03
Looking forward to constructive responses and guidelines
Kind regards,
A Beginner-level scholar
Your MEDMOD index is significant which fulfills your hypothesis of moderate mediation, in your case as you mention higher level of moderator reduces the effect so a negative effect otherwise would be a positive sign of MEDMOD
Dear Cristian Ramos-Vera , thanks for answering. I understood well about a negative effect is a positive sign of MEDMOD. Indeed, it supported my moderate mediation hypothesis, and demonstrated expected slops. Meanwhile, my concern for indirect effect and those three levels of moderation are still unclear to me. I would appreciate, if you guide me for that query as well. Thanks!
low :mean moderator-1desviation standar
mean moderator
high :mean moderator+1 desviation standar
Okay, great!
Does that mean low-level moderation has negative effect (-1 desviation stander), and it reduces the indirect effect. Right?
My hypothesis is;
Hypothesis 1a: TL moderates the indirect relationship between PB and Performance via JC, such that the relationship is amplified when TL is low.
Values at three levels of moderation are as follows;
Low = .45 Effect p< 0.00
Mean = .23 Effect p< 0.01
High = .16 Effect p< 0.03
Note: All values are positive (no subtraction sign) in analysis.
Looking forward to further guidance- Thanks!
- Badges
- Science topic
- Similar topics
- Computer Science
- Program