true for smaller sample sizes, but for large normally-distributed sets, such as “all people,” the mean and median will be pretty much identical. when you take a larger sample, it’s more likely that an outlier on one end will be balanced out by a similar outlier on the other end. e.g. when you measure the height of 50 people, you might happen to end up with one person who’s 6’8" and have to account for that, but when you measure the height of 50,000 people, you’re more likely to also have someone who’s 4’4"
Average is sensitive to outliers, it is possible more than half is below it (or above).
You want median.
I was paraphrasing a George Carlin joke:
true for smaller sample sizes, but for large normally-distributed sets, such as “all people,” the mean and median will be pretty much identical. when you take a larger sample, it’s more likely that an outlier on one end will be balanced out by a similar outlier on the other end. e.g. when you measure the height of 50 people, you might happen to end up with one person who’s 6’8" and have to account for that, but when you measure the height of 50,000 people, you’re more likely to also have someone who’s 4’4"