There are many factors that can affect testing results in a small child, beyond their actual cognitive ability, including fatigue, inattention, confusion about directions, anxiety, reluctance to guess, inexperience with timed tasks, relative personal weaknesses in fine motor dexterity, to name only a few. It is also not unheard of for the results of either the NNAT or the RIAS to be a bit different from those of other instruments, but this does seem to be on the higher end of the range for discrepancies between valid administrations.

For any re-test under your circumstances, the WISC-V would typically have been the recommendation (even if you believe the RIAS-2 was not a valid administration, and even without the test/re-test time), as, although the RIAS-2 has a bit more to it than the NNAT3, they are both brief/screening instruments, recommended as either part of multi-stage assessment, or for reevaluations, when a comprehensive evaluation has been completed previously. The WISC-V, on the other hand, is on the short-list of gold standard cognitive assessments for school-age children.

In any event, I typically do not recommend that a child--even a young child--who has already completed two standardized assessments engage in any kind of test prep. Sometimes this amplifies any anxiety they may have regarding the testing process, which can further depress performance. Although Qbitz can be fun for many children, so by all means play, if she enjoys it!
...pronounced like the long vowel and first letter of the alphabet...