There is a lot of hocus-pocus involved in both methods. Neither method will reveal the presence or absence of sinkholes, although it might give you supportive information.
I have looked at hundreds of geo reports with GPR, and have only seen 2 or 3 with Electric Resistivity. The GPR does not appear to be accurate at any depth over a few feet. The ER images, when I have been able to view them in color, were much more definitive in showing anomalies that may indicate ravelling, loose soils, or clay layers. I would say that Electric Resistivity is more promising than GPR. On the hundreds of reports I have seen, the GPR report does not even pick out the iron or copper water pipes going into and out of the home. This is in Florida, where the soil is mostly sand, with some areas having clay layers. There is a clay confining layer over the limerock in much of the state.
In no case is either of these methods (or any other method) capable of definitely locating a "sinkhole". They very seldom even find a definite ravelled zone. Both methods are limited in the depth to which they are accurate, with 20 feet as a practical maximum.
The method of SPT borings is the most revealing about subsurface conditions. Most geotech reports use only 2 SPTs, 2 test pits, 5 hand auger borings, and 8 to 12 GPR transects. There should be more borings than just 2 SPTs. I saw one report recently with 3 SPTs, 2 of which found limerock at 27 ft, and the third went 95 in sand with no limerock being found. Most of the houses that we have underpinned can have at least a 20 to 30 foot variation in limerock depths, with 60 foot variations being very common.
I have been told by geotech engrs that the only way to really understand the GPR is to be there reading it as it is moved along the soil surface. But even with that, their anomalies are not too accurate. Many times we have grouted where a GPR report says there are loose soils, but it will take no grout. Other areas on the same lot will take a lot more grout ( 1 or 2 CY). (When you hear about a property taking 100, 200, or 700 (!!??) cubic yards of grout, it has been pumped into the Aquifer in the limerock.)
I can give you the name of a soil engineer here in Florida who has written some papers on ER vs GPR, he would be better able to give you advice. There are different voltage and milliamps and frequency combinations that are probably important.
Janine Pardee PE
Structural/civil and mechanical
Orlando, FL