Literature shows that HR practices are negatively related to stress or turnover intention. I have organisational justice as a mediator between them. But my pearson corelation results shows that r=.265 for stress and r= .163 for turn over intention (TOI) which is a weak relationship. Can I still run mediation of org. justice between hr practices and stress/TOI . Is it necessary that my dependent and independent variable should have negative corelation.
Perhaps the error was originally due to the way by which the data has entered into statistical programs for analysis
It is normal to have negative relationships between study variables, although your expectations are on the positive side
I thank that these findings may be normal in some situation. You should take in consideration that there others variables out your hypothesized model may have influencing.
If your results are do not match the literature then you need to study more literature or perhaps increase your sample size. Even if you get opposing results it does not matter as long as you can justify properly in your discussion.
For mediation the necessary condition is that your independent variable and mediator should have significant relationship. If this condition is fulfilled you can go ahead with mediation. Search Baron and Kenny for help in mediation analysis and if you are using PROCESS search for papers by Andrew F. Hayes.
Good Luck
- Badges
- Science topic
- Similar topics
- Psychology
- Organizational Psychology