Psychology Statistics Tuition, SPSS Tuition in Singapore

There are two ways you can write mediation report in APA format.

1. Baron & Kenny (1986):

In this analysis, IV is family support, DV is current salary and MV is beginning salary. Following Baron and Kenny’s (1986) mediation analysis, a complete mediation happens when a) IV significantly predicts DV b) IV significantly predicts MV (path a is significant), c) MV significantly predicts DV, (path b is significant) and d) when mediator enters the IV-DV relationship, the total effect reduces significantly to non-significant. If the direct effect does not reduce significantly to non-significant, mediation only happens partially. In addition, the indirect effect for the mediational analysis has to be significant, and the Sobel test demonstrates that the value of indirect effect is greater than zero (z > 1.96). 

The results showed that there was a significant total effect between family support and currently salary (B = 1.86, p <.001), and path a (i.e., family support on beginning of salary) (B = 0.99, p <.001) and path b (i.e., beginning of salary on current salary) (B = 2.43, p <.001) were both significant. Finally, when beginning of salary entered the relationship between family support and currently salary, the direct effect (B = 0.52, p <.001) was significant. In addition, the Sobel test for the indirect effect is z = 24.13, p <.001; therefore, it was concluded that a partial mediation occurred between family support on current salary via beginning of salary.

Hayes Process (2013):

Following Hayes’ (2013) Macro Process via bootstrapping method, to consider a mediator has mediational effect when (1) the indirect effect (IE) of family support on current salary via beginning salary (i.e., IE = path a x path b; a = the effect of family support on the mediator of beginning salary, b = the effect of beginning salary on the current salary) and (2) the bias corrected 95% CI around the IE from 5000 bootstrap re-samples. We accepted the IE as statistically significant only if its bias corrected 95% CI excluded zero.

The results showed that there was a significant total effect between family support and currently salary (B = 1.86, p <.001), and path a (i.e., family support on beginning of salary) (B = 0.99, p <.001) and path b (i.e., beginning of salary on current salary) (B = 2.43, p <.001) were both significant. Finally, when beginning of salary entered the relationship between family support and currently salary, the direct effect (B = 0.52, p <.001) was significant. In addition, the bias corrected 95% CI is 1.16, and CI95% = 0.92 to 1.45 which excluded zero. Hence, beginning of salary is considered as a mediator for family support on current salary. 

Need help in your psychology statistics, SPSS or research report? SMS 97587925 or enquiry@starcresto.com