Mediate macro output categorical IV

#1
Hi,

I've come across similar threats but none have been answered unfortunately. For my bachelor thesis I have performed a regression analysis having coded my categorical IV into two dummies. Mediator and DV are both continuous. Only one of my IV conditions relates significantly to my mediator. I was told to use the Sobel test to determine the significance of the mediation effect, but haven't worked with this test before during my classes so it's completely new to me. Didn't take long to realize I can't enter the values from both dummies however. How can I solve this, or should I use another test? Or should I not test significance at all seeing only one of the dummies correlates with the M?
My thesis instructor is being no help so any tips would be much appreciated!

---UPDATE---

I have run the mediate macro as provided by AF Hayes but am struggling with the output. What significance levels should I be looking at, and what do the LLCI and ULCI mean? Also, my thesis instructor is still maintaining I should run an analyses for the IV as a whole which I am still convinced is impossible. Should I conclude there is no mediation because the effect of X on M is only significant on one condition? The "indirect effects through support _" part of the outcome isn't giving me any p levels. Suggestions?

MATRIX procedure:

*************** MEDIATE Procedure for SPSS Release 050213 ****************

Written by Andrew F. Hayes, Ph.D. http://www.afhayes.com

**************************************************************************

VARIABLES IN THE FULL MODEL:
Y = appmot_s
M1 = support_
X = C1
C2

**************************************************************************

OUTCOME VARIABLE:
support_

MODEL SUMMARY
R R-sq Adj R-sq F df1 df2 p
,3341 ,1116 ,0914 5,5280 2,0000 88,0000 ,0055

MODEL COEFFICIENTS
Coeff. s.e. t p
Constant 4,6229 ,2013 22,9670 ,0000
C1 -,7783 ,2870 -2,7117 ,0081
C2 -,8637 ,2870 -3,0091 ,0034

**************************************************************************

OUTCOME VARIABLE:
appmot_s

MODEL SUMMARY
R R-sq adj R-sq F df1 df2 p
,6889 ,4745 ,4564 26,1905 3,0000 87,0000 ,0000

MODEL COEFFICIENTS
Coeff. s.e. t p
Constant ,8765 ,3646 2,4039 ,0183
support_ ,5687 ,0730 7,7880 ,0000
C1 -,2406 ,2046 -1,1755 ,2430
C2 -,2587 ,2065 -1,2529 ,2136

TEST OF HOMOGENEITY OF REGRESSION (X*M INTERACTION)
R-sq F df1 df2 p
support_ ,0036 ,2957 2,0000 85,0000 ,7448

OMNIBUS TEST OF DIRECT EFFECT
R-sq F df1 df2 p
,0116 ,9639 2,0000 87,0000 ,3855

**************************************************************************

INDIRECT EFFECT(S) THROUGH:
support_


Effect SE(boot) LLCI ULCI
C1 -,4426 ,1862 -,8533 -,1298
C2 -,4911 ,1995 -,9243 -,1464
OMNIBUS ,0520 ,0409 -,0020 ,1380

----------