FWIW, I don't think false high scores are possible. Many issues can interfere with testing and create false lows, but not false high.
This becomes a much bigger topic, but I have to say that I don't agree with this. Especially in young children, exposure makes a big difference on IQ tests. Kids who have enriched environments are much more likely to test highly on tests of vocabulary, knowledge of social norms and general knowledge, puzzles, etc. All of these make up parts of common well regarded IQ tests such as the WISC. It doesn't require "hot housing" per se with the intent of inflating scores. Likewise, scores can be depressed by un-enriched environments.
I believe that is why adoption studies show that adoptees' IQ scores are more similar to those of their adoptive parents in childhood and regress or inflate to be more in line with those of their biological parents the older they get.
All of this is not to say that the IQ scores are necessarily wrong in the OP's instance. Like someone else mentioned, there could be a dual exceptionality such as a learning disability at play.
