In the attached model, sobel test is being used to test for mediation effects. one of my Hypotheses says: Job satisfaction mediates the relationship between HRM practices and turnover intentions. Sobel statistic suggests that job satisfaction does not bediate between HRM practices and turnover. I am afraid this would be misleading since job satisfaction mediates between HRM and the second mediator, organisational commitment.
But logic tells me that there is mediation from HRM to Job satisfaction to organisational commitment to turnover intentions.
How do I get a statistic to say that.