Wouldnt mean salary be a better representation of the public then median?
In: Other
In addition to what others have said about outliers, in context of salary, the difference from a “medium” salary to a very high salary is much different than to a low salary. This makes the mean likely to be higher than what people would usually think of as an average salary.
For example, if you have 10 people making the following:
$3,000
$25,000
$40,000
$40,000
$40,000
$75,000
$75,000
$100,000
$200,000
$3,000,000
The mean is $359,800 – it doesn’t accurately reflect anyone in the group. Even if you took the second and third highest and made them zero earners, the highest has skewed it so far that the mean is still $329,800. It took just one very high earner to shift the mean more than the zeros can offset.
The median on the other hand moves up and down a couple spots if you add really high or really low earners, but if there’s a consistent(ish) middle, it will still be roughly in the middle. And the highest earners won’t push it so far out of that range.
Averages are averages, they give us a simpler way to absorb data but can’t give a full picture of complex information. The types of numbers you’re working with can make mean or median better reflect the information you want to talk about. Sometimes outliers balance each other and the mean is useful. Sometimes they don’t.
Sometimes you don’t want either and modal is preferred. For example if you want to know for a totally random individual that you bump into on the street what is their most likely income bracket then what you actually care about is modal income. However for a variable like income it forces you to group income ranges which can be a bit arbitrary.
Genuinely why not use the mode and compare median and mode together for a more accurate picture?
Because if Bill Gates or Elon Musk walks into a bar, everyone in the bar, *on average,* is a millionaire. But this is not an accurate way of describing how rich most people in the bar actually are.
When you’re looking at this you really aren’t looking for the average salary, you’re looking for the average person’s salary which would point to median over mean.
No, because salary isn’t evenly distributed.
Imagine a group with a mean of $30K. Then add one millionaire. The mean would go way way up, but only one person is wealthier and the rest are just as poor as they used to be.
There are a lot of good answers here but I would add:
Salary is lower-bounded at but without any upper bound. As a result, the mean will naturally skew higher than the other measures of center (median and mode).
As a result, the mean is a poorer reflection of the actual circumstances under which people live than the other two in this instance.
Because of the massive and growing income inequality. The more fair and equal payments become the mean will be closer to the median value and otherwise deviate. When we mean the “average income” in common language, we are actually looking for a value that truly reflects the experiences of most people, not the strict statistical average of incomes.
Mean is when an oligarch makes a million and peasant is making one dollar and mean average between them is half a million and 50 cents. Not representative of anything.
Because the central limit theorem doesn’t really work for anything with money. I don’t know why.
Latest Answers