Twice exceptionality might. If the original test was given when the child was young, the numbers might have been inflated or deflated due to environmental factors (or anxiety or lack of comfort with strangers/the tester if the first IQ score was significantly lower).
Scores btwn the WISC-III and IV are also not going to be exactly spot on. I understand that kids who were tested on both versions often got a lower # on the WISC-IV, not to the extent you have here, but somewhat lower. I don't know if this is due to the Flynn Effect or just the later test scoring lower, or what.
