Doesn’t using mode just make a lot more sense? You are much more likely to be the mode than you are likely to be average in the mean class
[Originally said average in the title, fixed thanks to jbrains]
Doesn’t using mode just make a lot more sense? You are much more likely to be the mode than you are likely to be average in the mean class
[Originally said average in the title, fixed thanks to jbrains]
Mode is easily be screwed up by data distribution, it’s used for almost nothing.
1, 2, 3, 4, 5, 45, 98, 100, 100
Mode = 100 Mean = 39.8 Median = 45
Mean is used because it handles most data-sets well, it just gets screwed up a lot when there are big outliers. That’s when we use Median instead.
For example we use Median a lot for household incomes, there’s a lot of data points, but there are some ridiculously wealthy people and some people that make absolutely nothing. Mode doesn’t really work in this situation, because it would likely just be 0.
1, 2, 3, 4, 5, 45, 98, 100, 100 => 100
1, 2, 3, 3, 5, 45, 98, 99, 100 => 3
1, 2.999, 3, 4, 5, 45, 98, 99.999, 100 => ?
It will only work on datasets where you have few very distincts values but also lots of values.
I guess we don’t like it eather because shitty headlines would have a field day if people even knew “mode” might be about in statistics (or am I thinking too french?). Thank god for that naming BTW.