I don't know which test was used, and there are several, but usually when a larger sample is used in any quantitative analysis it is because the product being measured is too small to be measured accurately. By increasing the amount of the item being measured it increases the accuracy somewhat. That may or may not have been the reason in this case.
