This is a debatable notion, indeed one that I believe has been debated on this forum before. The question is this: If the code gives you a specific minimum VA for a specific type of load, should we presume that the code authors knew it was going to be a continuous load, and therefore already added the extra 25% before they told us what number to use? In other words, was the intended load for the sign circuit actually 960 VA, but they multiplied that by 125% and gave us the 1200 VA result as the value to use? It is possible that the test question's author took the opposite point of view, and saw the 1200 as the non-continuous value, and added the extra 25% to get a final answer of 1500. To me, that sounds like the most likely explanation.
For my part, when I see a value in an NEC table, I do not apply a 125% factor for continuous loads. For example, if I design a lighting system, including selection of light fixtures and lamps, I will calculate the lighting load by putting the actual lamp loads on a panel schedule. I will apply the 125% factor in that calculatio. By contrast, if I am using 3VA per square foot as the amount of lighting load assigned to a service calculation, I will not multiply the 125% factor to that part of the calculation. I believe the fact that lights can be on more than 3 hours is already bullt into the 3VA/SF value that I pulled from the table.
