Sometimes data does have a huge difference and mean is still the best way to find the central point. For example. Perhaps you have a millionaire, someone making $30,000 and a lot of people spread out all over (and evenly) in between. How would this need to change for median to be a better measure?