If you are talking about arrays of boolean values, please bear in mind that a boolean in LV occupies 8 bits. I remember reading that this was changed over previous versions (LV 4 to LV5?) for compatibility and simplification reasons.
If you need to generate a 32-bit number from 32 booleans, there is a function to convert a boolean array to a scalar. Simply "sticking" 32 booleans together and then assuming you have the equivalent of a 32 bit number is incorrect. Hope this helps Shane.