i am facing a problem in which IVs and mediator are negative behaviors. correlation value is positive which shows that IVs have strong association with mediator, which is understandable because both are negative behavioral aspects. but problem starts when i introduce a positive behavioral aspect as a dependent variable, and correlation value for mediator and dependent variable is also positive. All regression coefficients are also positive.
should i reverse code the values for mediator? or is there any other solution
kindly have a look on attached image.
X= 2nd order independent variable
M= mediator
Y= Dependent Variable
answer with reference will be highly appreciated