Is this possible? The negative moderation effect is small (both effects are significant). The mediation simultaneously included 6 other variables, another 2 of which also mediated the relationship between SK and SU. I just want to know whether this is statistically possible, but I also need some help to interpret what it means. Thank you.