That's the big problem for researchers; we have plenty of evidence that the ballads were sometimes edited/recreated/made up completely by the editors like Scott, but we have no way of knowing to what extent. You can get a good idea by reading carefully Child's headnotes, particularly in the first 3 volumes.

There's an awful lot of material in there that has no evidence whatsoever for oral tradition, and quite often the supposed oral extant versions come from only a few suspect sources.