So if "mode" occurs most frequently, it will have the biggest impact on the mean/average, yes?
Clearly 1 had a bigger impact than any of the other numbers.
You'd think so. And it's true - except. Say we have a list of 53 values:
6, 6, 6, 601, 602, 603, 604, 605, 606, ... ,650.
The mean (average) is close to 600. In many real problems you might get data kinda similar to this (readings from a sensor). In this case, our mode is not very descriptive of the data at all. We might be tempted to just call them 'outliers' and throw them away. But we have to be careful. It's often a judgement call - is it real data or an equipment malfunction (or maybe just an quirk of the equipment or procedure).
MG: 33%
WW: 61% (I'm guessing, okay?! but not with regard to my own mark! I did get that once, on a math test!)
wow: 66%
wwh: 75%
dodyskin: 75%
consuela: 75%
etaoin (hope I spelled that right): 77%
goat: 88%
stales: 95%
inselpeter: 98%
Let me line the data up so I can read it:
33 61 66 75 75 75 77 88 95 98
If I punched into my calculator correctly, the average is 74.3%
(743/10). Note that 74.3% is not a value in your list. The mode
is 75 and the median is also 75. Frankly, I don't use mode very
much (I don't know why - it's just never come up). Well, I guess
I do sorta use it implicitly in histograms, but I don't normally
(ahem) think about it. Then again, I'm not a professional statistician.
Btw, you can remember etaoin's name, which you did spell correctly, as the most common letters in the English language in descending order, etaoin shrdlu, which I believe Bill, at least, already hinted at.
k
All that and I miscounted. I modified my example and forgot to recount. Say 53 (or so items, with a few duplications even, but no n-tuples with n greater than 2).
