M 2 = .01 ). However, in order to determine support for Hypotheses 1 and 2, we graphed our interactions. Fig. 1 provides a graphical representation of the interaction with job satisfaction as the dependent variable. As seen in this figure, the greatest rate of change for the job satisfaction line slope occurred for those respondents who reported low empowerment. Conversely, there was little change in the job satisfaction line slope for those respondents who reported high empowerment. Additionally, the positive
relationship between LMX and job satisfaction was strongest when empowerment was lowest. Results from simple slopes tests, shown at the bottom ofTable 3, confirm this interpretation. Taken together, these results provide support for Hypothesis 1.
Fig. 2 provides a graphical depiction of the LMX*empowerment interaction with turnover intentions as the outcome. As seen in this figure, the rate of change for the turnover intentions line slope is strongest for those respondents who reported low empowerment, while the rate of change for the turnover intentions line slope for high empowerment was less strong. Interpreting this interaction, we again see the moderating influence of empowerment on the LMX quality-outcomes relationship. In this case, the respondents reporting low empowerment coupled with low quality LMX relationships reported the highest turnover intentions. The result from our simple slopes test on this interaction, which is at the bottom of Table 3, supports this explanation. These results support Hypothesis 2.