NWEA has a PDF documenting a study to generate correlation scores. http://www.nwea.org/sites/www.nwea.org/files/resources/Iowa%20Linking%20Study%20August%202010.pdf

With correlations around .8 for 4th grade math and .75 for 4th grade reading, then I think it is fair to say highly likely to score at 99 on ITBS given a 99 on MAP.

But 99 is a fickle number as it represents all the top scores beyond it. So, a number functionally further out like a 99 that represents a 99.99 is incredibly likely that the comparison test will also be above 99. Whereas a score that hits right at 99, is more like a 50/50 chance of another 99 but a highly to get a 98+. Stuff like that.

The other problem with a full test to test correlation is one test may include a single subtest with really poor correlation. Example would be if one test includes a handwriting subtest and the other doesn't. Then if your kid has a particular weakness in handwriting, that could have a fairly dramatic impact in the tight range where gifted scores exist.