Unshielded
Wire for pitot heat power leads does not need to be shielded. Shielded wire is usually used to protect the wire from external emf (electro motive force or electro magnetic fields) for very low signal level wires (microphone, low level audio, sensor signals, etc), and to keep high level noisy (ie pulsed voltages such as strobe light drive voltages, mgneto p-leads, etc.) contained within the wire (in this case it doesn't so much contain the magnetic field as it does provide a capacitive "short" to transigent (quickly changing) voltages).
High current wires such as pitot heat, landing lights, or any main power wires are fairly constant current DC voltages and therefore don't generate a lot of noise, so they are not normally shielded. However, the high current generates a magnetic field around the wire. The higher the current, the more magnetic field (this is how hall effect current sensors work and, for that matter the basis of how electric motors and generators work -- so it's a good thing).
Usually this does not cause problems unless the magnetic field encroaches upon something (like a compass or magnetometer) that out of necessity is sensitive to magnetic fields. If the wires are kept an appropriate distance (a foot or more, if possible, at currents we're talking about) from the compass (magnetometer, etc) we should be OK. Unless, like mentioned above, there are loops in the wire, which could amplify the magnetism (a little difficult to explain).
Also, if there is a ground return wire from the pitot heat to a central ground block, it would carry the same current as the power side and would generate the same magnetic characteristics as the power wire. Usually, high power grounds are grounded to the aircraft frame (in aluminum a/c such as RV's) near the load (pitot tube, landing light, etc.) rather than returned to a central ground, but since this is a homebuilt a/c there is an almost equal chance it is returned. So you need to check the ground side of the pitot heat as well as the power side. If it's grounded at or near the pitot tube - no problem.
Your problem is almost certainly due to wire routing issues. A lot of panel mounted compasses have this or similar problems. If you can't get to the root of the problem, you may need to relocate the compass.
I hope this helps and doesn't confuse you more.