If factors are only rounded up, then the estimate is an overestimate. If factors are only rounded down, then the estimate is an underestimate. When some factors are rounded up and some are rounded down, it is harder to tell whether the estimate is an overestimate or an underestimate.
